The impact of macro- and micro-steatosis on the outcomes of patients who undergo liver transplant: analysis of the UNOS-STAR database

Abstract

Background & Aims:

The presence of steatosis in a donor liver and its relation to post-transplantation outcomes are not well-defined. This study evaluates the effect of the presence and severity of micro and macrosteatosis of a donor graft on post-transplantation outcomes.

Methods:

The UNOS-STAR registry (2005–2019) was used to select patients who received a liver transplant graft with hepatic steatosis. The study cohort was stratified by the presence macrovesciular or microvesicular steatosis, and further stratified by histologic grade of steatosis. The primary endpoints of all-cause mortality and graft failure were compared using sequential Cox regression analysis. Analysis of specific causes of mortality were further performed.

Results:

There were 9,184 with no macrosteatosis (control), 150 with grade 3 macrosteatosis, 822 with grade 2 macrosteatosis, and 12,585 with grade 1 macrosteatosis. There were 10,320 without microsteatosis (control), 478 with grade 3 microsteatosis, 1,539 with grade 2 microsteatosis, and 10,404 with grade 1 microsteatosis. There was no significant difference in all-cause mortality or graft failure among recipients who received a donor organ with any evidence of macro or microsteatosis, compared to those receiving non-steatotic grafts. There was increased mortality due to biliary and infectious causes among recipients with a donor organ of grade 3 macrosteatosis and grade 3 microsteatosis, and increased mortality due to cardiac arrest, emboli, graft infection, disseminated intravascular coagulation hemorrhage, and other vascular pathologies among recipients of a grade 2 macrosteatosis donor organ.

Conclusion:

This study shows no significant difference in all-cause mortality or graft failure among recipients who received a donor liver with any degree of micro or macrosteatosis. Further analysis identified increased mortality due to specific etiologies among recipients receiving donor organs with varying grades of macro and microsteatosis.

Lay Summary

This study looked at how fat deposits in donor livers, known as steatosis, might affect the success of liver transplantation. The study analyzed data from over 20,000 liver transplant cases from 2005 to 2019. The results showed that when recipients received livers with any level of fat deposits, there wasnť a significant difference in overall survival or graft failure compared to those who got livers without fat deposits. However, when they looked closer at specific causes of death, they found that recipients of livers with severe fat deposits (grade 3) faced higher risks of mortality from biliary and infectious complications following liver transplant. Additionally, those with moderately fatty livers (grade 2) had increased risks of death from cardiac arrest, clots, infection of the transplant, bleeding disorders, and other vascular problems.

Introduction

Among patients with advanced liver disease, liver transplantation (LT) is a definitive therapeutic option that reverses the symptoms of organ failure and portal hypertension (1). Long-term prognoses after organ allocation are often favorable, and improvements in short-term patient and graft survival have been especially pronounced (2–4). However, adverse LT outcomes still occur, and the risks for these undesirable events are dependent on recipient, graft, and donor-related variables (5–8). In particular, there has been a budding interest in understanding if graft steatosis has any bearing on LT recipient outcomes (9). There have been conflicting presentations on this topic, with some reviews observing increased mortality among recipients who received grafts with underlying steatosis, and other studies suggesting these histological changes of steatosis to be innocuous (10).

Given this conflicting evidence, further investigation is required to clarify the precise effects of graft microvesicular and macrovesicular steatosis on recipient and graft survival following LT. As the recipient pool continues to enlarge and the available organ supply becomes further outstripped, UNOS-distribution mechanics and screening processes have adapted to meet the growing organ demand (11). As policy changes search for a solution to meet the supply-demand discrepancy and the literature advocates for expansion to the donor pool, the effects of graft steatosis on both recipient and graft outcomes have become more relevant (12,13).

In this study, we use the United Network for Organ Sharing (UNOS) Standard Transplant and Analysis Research (STAR) database to evaluate LT recipients whose grafts are from donors with either micro or macro-steatosis, in order to capture the effects of steatosis on the post-LT outcomes. Furthermore, we stratify the severity of steatosis to differentially classify the prognostic value of this histological distortion on recipient and graft outcomes.

Methods

Database and Study Cohort

The UNOS STAR registry compiles patient outcome and follow-up information for American transplant patients. Our study utilized data from 2005 to 2019 from the UNOS STAR database. To guarantee confidentiality, all patient data was masked along with data-use agreements and safety mechanisms. This study was supported by the Health Resources and Services Administration Contract 234–2005-370011C. Views and opinions of this manuscript are reflected upon the authors and do not represent any stance of any governmental body.

Study Population and Variables

A total of 99,987 patients in the UNOS database received a liver transplant between 2005 and 2019. Exclusion criteria were applied, which were as follows: cases without data pertaining to the exposure (hepatic micro and macrosteatosis) (n=71,647), lost to follow up (n=773), those that had a history of prior LT (n=1,014), under the age of 18 (n=324), had a non-heart-beating organ donation (n= 846), underwent partial LT (n=1,426), and that underwent retransplantation (n=1,154). This left a final population of 22,741patients, which was then stratified using the donor steatosis severity contained in UNOS. Patients who underwent re-transplantation were excluded from the analysis in order to mitigate bias as retransplantation has been associated with poorer outcomes compared to primary transplantation (14, 15). Figure 1 delineates the patient selection process underlying the final cohort selection.

Figure 1.

This figure shows the patient selection process of this study

Degree of hepatic steatosis in the donor graft was obtained from the UNOS registry. UNOS requires participating institutions to stain hepatopathology slides through standardized protocols with hematoxylin and eosin, as well as fat-specific methods (Red-oil-O or Sudan red) in order to identify fatty deposits in the parenchyma and quantify the area of change relative to the total examined parenchyma. Degree and grade of macro versus microsteatosis is then interpreted per the methods of each participating institution and pathologist. Steatosis grading was performed using cross-sectional hepatic histology and various threshold percentages, in which 0 to 5% indicated stage 0, 3 to 33% indicated stage 1, 33 to 66% indicated stage 2, and 66 to 100% indicated stage 3. The escalating severities of steatosis were used to stratify the sample and to compare the groups with the study endpoints, which included the primary endpoints of all-cause deaths and graft failure as well as secondary system-specific causes of death identified in the database. The specific etiologies of death in the secondary analysis included mortality due to biliary complication, cardiac arrest, pulmonary embolism, hemorrhage, graft failure, as well as general neurological, renal, infectious, respiratory, and vascular causes.

In addition to the primary comparisons, a strata-specific approach was employed to subdivide the sample into those who received donor livers with mixed micro and macrovesicular steatosis (compared to those with pure micro and macrosteatosis). Covariates included recipient demographics (age, gender, race, and body mass index), comorbidities (hepatitis B, hepatitis C, alcoholic liver disease, diabetes), hepatic variables (ascites, encephalopathy, history of transjugular intrahepatic portosystemic shunt (TIPS), and model for end-stage liver disease (MELD) scores), medications (mycophenolate mofetil, cyclosporine, tacrolimus, sirolimus, steroids), biomarkers (albumin, creatinine, INR, total bilirubin), life support variables (artificial liver devices, utilization of inotropic agents, ICU admissions, ventilator support), donor demographics (age, gender, race, body mass index), and donor biomarkers (donor creatinine and donor total bilirubin).

Statistical Analysis

Baseline characteristics were evaluated between the steatotic cohorts. To determine the significance of nominal variables, parametric factors, and non-parametric factors, the Fisher’s or Chi Squared test, Student’s t test, and Whitney-U tests were used, respectively. Primary and secondary outcomes were analyzed with iterative Cox regression. Each successive model included more covariates, as well as the associated adjusted hazard ratio, 95% confidence interval, and p-value. Model 1 included recipient demographics. Model 2 included model 1 covariates and recipient comorbidities. Model 3 included model 2 covariates, recipient liver status, and laboratory markers. Model 4 included model 3 covariates and donor characteristics. For each strata-outcome relationship, we reported the incidence rates in units per 1000 person-years. We further analyzed our primary outcomes with cumulative hazard analyses with log-rank testing to determine the significance of the analyses. Competing risk regression was also performed according to the procedure outlined by Fine and Gray (16). Statistical analysis was performed with RStudio version 1.2.5042 with R code version 3.6.3.

Results

Baseline Characteristics

Table 1A details the baseline characteristics of our study's patient population. Tables 1A and 1B contain information from liver transplant recipients who had donors with macro-steatosis and micro-steatosis respectively. Patients were divided into groups of grade 3, grade 2, grade 1 or no donor steatosis with additional covariates stratifying them. Significant findings were identified with p-values of <.05 being significant.

Table 1A:

Baseline Characteristics of Liver Transplant Recipients Stratified Using Grade of Macrovesicular Steatosis

	No Steatosis			Grade 1 Macrovesicular Steatosis					Grade 2 Macrovesicular Steatosis					Grade 3 Macrovesicular Steatosis
Variable	n = 9184	42.19	%	n = 12585	57.81	%	p-value		n = 822	8.22	%	p-value		n = 150	1.61	%	p-value
Recipient Demographics
Age (years)	56.10	± 9.57	years	56.20	± 9.46	years	0.28		56.50	± 9.47	years	0.27		55.80	± 8.51	years	0.44
Gender (%)	6208	67.60	%	8972	71.29	%	< 0.001	***	602	73.24	%	0.001	**	107	71.33	%	0.38
Race							0.14					0.52					0.33	†
White (%)	6700	72.95	%	9306	73.95	%			591	71.90	%			117	78.00	%
Black (%)	811	8.83	%	1045	8.30	%			65	7.91	%			14	9.33	%
Hispanic (%)	1183	12.88	%	1520	12.08	%			122	14.84	%			12	8.00	%
Asian (%)	358	3.90	%	507	4.03	%			31	3.77	%			4	2.67	%
Other (%)	132	1.44	%	207	1.64	%			13	1.58	%			3	2.00	%
BMI (kg/m²)	28.90	± 5.71	kg/m²	29.00	± 5.71	kg/m²	0.34		29.00	± 5.70	kg/m²	0.66		28.60	± 5.63	kg/m²	0.69
Comorbidities
Hepatitis B (%)	507	5.52	%	622	4.94	%	0.06		41	4.99	%	0.57		5	3.33	%	0.32
Hepatitis C (%)	4049	44.09	%	5394	42.86	%	0.07		317	38.56	%	0.003	**	64	42.67	%	0.79
Alcoholic Liver Disease (%)	2222	24.19	%	3374	26.81	%	< 0.001	***	232	28.22	%	0.01	*	41	27.33	%	0.43
Diabetes (%)	2093	22.79	%	2928	23.27	%	0.42		191	23.24	%	0.80		34	22.67	%	1.00
Hepatic Variables
Ascites							0.03	*				0.74					0.38
Absent (%)	2328	25.35	%	3313	26.32	%			216	26.28	%			31	20.67	%
Slight (%)	4341	47.27	%	5726	45.50	%			390	47.45	%			73	48.67	%
Moderate (%)	2515	27.38	%	3546	28.18	%			216	26.28	%			46	30.67	%
Encephalopathy							0.39					0.64					0.29
None (%)	3505	38.16	%	4756	37.79	%			326	39.66	%			48	32.00	%
1 – 2 (%)	4779	52.04	%	6525	51.85	%			421	51.22	%			87	58.00	%
3 – 4 (%)	900	9.80	%	1304	10.36	%			75	9.12	%			15	10.00	%
MELD Scores (unit)	20.10	± 9.62	unit	20.10	± 9.63	unit	0.87		18.40	± 8.42	unit	< 0.001	***	19.90	± 8.66	unit	0.93
Medications
Mycophenolate Mofetil (%)	7430	80.90	%	10074	80.05	%	0.12		654	79.56	%	0.37		131	87.33	%	0.06
Cyclosporine (%)	458	4.99	%	627	4.98	%	1.00		24	2.92	%	0.01	*	6	4.00	%	0.72
Tacrolimus (%)	8398	91.44	%	11428	90.81	%	0.11		735	89.42	%	0.06		137	91.33	%	1.00
Sirolimus (%)	223	2.43	%	275	2.19	%	0.25		9	1.09	%	0.02	*	4	2.67	%	0.79	†
Steroids (%)	8250	89.83	%	11077	88.02	%	< 0.001	***	706	85.89	%	< 0.001	***	130	86.67	%	0.26
Biomarkers
Albumin (mg/dL)	3.06	± 0.71	mg/dL	3.08	± 0.72	mg/dL	0.16		3.08	± 0.70	mg/dL	0.77		2.98	± 0.71	mg/dL	0.14
Creatinine (mg/dL)	1.31	± 0.96	mg/dL	1.34	± 0.96	mg/dL	0.006	**	1.18	± 0.75	mg/dL	< 0.001	***	1.25	± 0.86	mg/dL	0.82
INR (unit)	1.87	± 1.04	unit	1.87	± 1.07	unit	0.90		1.74	± 1.03	unit	< 0.001	***	1.87	± 1.28	unit	0.77
Total Bilirubin (mg/dL)	7.53	± 10.10	mg/dL	7.26	± 9.73	mg/dL	0.06		6.21	± 8.19	mg/dL	0.003	**	7.27	± 9.85	mg/dL	0.46
Life Support Variables
Artificial Liver Devices (%)	2	0.02	%	3	0.02	%	1.00	†	0	0.00	%	1.00	†	0	0.00	%	1.00	†
Primary Inotropic Agent							< 0.001	***				0.02	*				< 0.001	***
Dobutamine (%)	189	2.06	%	283	2.25	%			26	3.16	%			6	4.00	%
Dopamine (%)	1459	15.89	%	1849	14.69	%			104	12.65	%			19	12.67	%
Epinephrine (%)	98	1.07	%	144	1.14	%			8	0.97	%			6	4.00	%
Levophed (%)	1702	18.53	%	2245	17.84	%			137	16.67	%			20	13.33	%
Neosynephrine (%)	1459	15.89	%	2347	18.65	%			151	18.37	%			38	25.33	%
None (%)	4092	44.56	%	5452	43.32	%			376	45.74	%			52	34.67	%
Other (%)	185	2.01	%	265	2.11	%			20	2.43	%			9	6.00	%
Secondary Inotropic Agent							0.43					0.56					0.41	†
Dobutamine (%)	62	0.68	%	65	0.52	%			5	0.61	%			0	0.00	%
Dopamine (%)	90	0.98	%	124	0.99	%			7	0.85	%			2	1.33	%
Epinephrine (%)	60	0.65	%	77	0.61	%			6	0.73	%			2	1.33	%
Levophed (%)	372	4.05	%	488	3.88	%			22	2.68	%			4	2.67	%
Neosynephrine (%)	431	4.69	%	653	5.19	%			45	5.47	%			11	7.33	%
None (%)	8011	87.23	%	10977	87.22	%			723	87.96	%			128	85.33	%
Other (%)	158	1.72	%	201	1.60	%			14	1.70	%			3	2.00	%
Ternary Inotropic Agent							0.37					0.96	†				0.86	†
Dobutamine (%)	14	0.15	%	12	0.10	%			2	0.24	%			0	0.00	%
Dopamine (%)	5	0.05	%	8	0.06	%			0	0.00	%			0	0.00	%
Epinephrine (%)	21	0.23	%	18	0.14	%			1	0.12	%			0	0.00	%
Levophed (%)	45	0.49	%	58	0.46	%			3	0.36	%			1	0.67	%
Neosynephrine (%)	50	0.54	%	54	0.43	%			3	0.36	%			1	0.67	%
None (%)	9011	98.12	%	12370	98.29	%			810	98.54	%			148	98.67	%
Other (%)	38	0.41	%	65	0.52	%			3	0.36	%			0	0.00	%
ICU Admission							0.55					0.02	*				0.06
ICU (%)	867	9.44	%	1157	9.19	%			56	6.81	%			7	4.67	%
No ICU (%)	8317	90.56	%	11428	90.81	%			766	93.19	%			143	95.33	%
Ventilator Support (%)	310	3.38	%	393	3.12	%	0.32		16	1.95	%	0.04	*	4	2.67	%	0.82	†
TIPS Procedure (%)	929	10.12	%	1233	9.80	%	0.45		84	10.22	%	0.97		18	12.00	%	0.53
Donor Demographics
Donor Age (years)	48.80	± 15.80	years	50.40	± 14.30	years	< 0.001	***	45.50	± 13.50	years	< 0.001	***	46.20	± 14.90	years	0.05
Donor Gender (%)	4838	52.68	%	6880	54.67	%	0.004	**	448	54.50	%	0.33		82	54.67	%	0.69
Donor Race							< 0.001	***				< 0.001	***				0.09	†
White (%)	5958	64.87	%	8615	68.45	%			559	68.00	%			108	72.00	%
Black (%)	1898	20.67	%	2058	16.35	%			71	8.64	%			22	14.67	%
Hispanic (%)	903	9.83	%	1400	11.12	%			165	20.07	%			18	12.00	%
Asian (%)	225	2.45	%	304	2.42	%			20	2.43	%			1	0.67	%
Other (%)	200	2.18	%	208	1.65	%			7	0.85	%			1	0.67	%
Donor BMI (kg/m²)	28.30	± 7.02	kg/m²	30.80	± 7.51	kg/m²	< 0.001	***	31.90	± 7.97	kg/m²	< 0.001	***	30.80	± 6.85	kg/m²	< 0.001	***
Donor Biomarkers
Donor Creatinine (mg/dL)	1.95	± 2.04	mg/dL	1.85	± 1.86	mg/dL	0.81		1.77	± 1.59	mg/dL	0.63		1.80	± 1.70	mg/dL	0.86
Donor Total Bilirubin (mg/dL)	0.90	± 0.95	mg/dL	0.89	± 0.88	mg/dL	0.003	**	0.90	± 0.72	mg/dL	0.01	*	0.88	± 0.73	mg/dL	0.66

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†Fisher's Test

Isolative sample indicates that patients had no alternative liver diagnoses except NASH; hence, patients with hepatitis B, C, alcoholic liver disease, and HCC were excluded

Table 1B:

Baseline Characteristics of Liver Transplant Recipients Stratified Using Grade of Microvesicular Steatosis

	No Steatosis			Grade 1 Microvesicular Steatosis					Grade 2 Microvesicular Steatosis					Grade 3 Microvesicular Steatosis
Variable	n = 10320	49.80	%	n = 10404	50.20	%	p-value		n = 1539	12.98	%	p-value		n = 478	4.43	%	p-value
Recipient Demographics
Age (years)	56.10	± 9.45	years	56.30	± 9.51	years	0.13		55.90	± 9.77	years	0.72		55.30	± 9.16	years	0.04	*
Gender (%)	7101	68.81	%	7344	70.59	%	0.006	**	1103	71.67	%	0.03	*	341	71.34	%	0.26
Race							0.24					0.08					0.48
White (%)	7634	73.97	%	7633	73.37	%			1102	71.60	%			345	72.18	%
Black (%)	870	8.43	%	884	8.50	%			132	8.58	%			49	10.25	%
Hispanic (%)	1255	12.16	%	1298	12.48	%			222	14.42	%			62	12.97	%
Asian (%)	384	3.72	%	436	4.19	%			63	4.09	%			17	3.56	%
Other (%)	177	1.72	%	153	1.47	%			20	1.30	%			5	1.05	%
BMI (kg/m²)	29.00	± 5.74	kg/m²	28.90	± 5.71	kg/m²	0.50		28.70	± 5.48	kg/m²	0.17		29.10	± 5.65	kg/m²	0.52
Comorbidities
Hepatitis B (%)	541	5.24	%	536	5.15	%	0.79		76	4.94	%	0.66		22	4.60	%	0.61
Hepatitis C (%)	4488	43.49	%	4480	43.06	%	0.54		646	41.98	%	0.28		210	43.93	%	0.89
Alcoholic Liver Disease (%)	2590	25.10	%	2762	26.55	%	0.02	*	389	25.28	%	0.90		128	26.78	%	0.44
Diabetes (%)	2379	23.05	%	2424	23.30	%	0.69		340	22.09	%	0.42		103	21.55	%	0.48
Hepatic Variables
Ascites							< 0.001	***				0.08					0.002	**
Absent (%)	2610	25.29	%	2744	26.37	%			421	27.36	%			113	23.64	%
Slight (%)	4933	47.80	%	4704	45.21	%			691	44.90	%			202	42.26	%
Moderate (%)	2777	26.91	%	2956	28.41	%			427	27.75	%			163	34.10	%
Encephalopathy							0.01	*				0.005	**				0.17
None (%)	3821	37.03	%	4025	38.69	%			628	40.81	%			161	33.68	%
1 – 2 (%)	5484	53.14	%	5318	51.11	%			750	48.73	%			260	54.39	%
3 – 4 (%)	1015	9.84	%	1061	10.20	%			161	10.46	%			57	11.92	%
MELD Scores (unit)	20.20	± 9.57	unit	20.10	± 9.65	unit	0.17		19.30	± 9.31	unit	< 0.001	***	19.20	± 9.04	unit	0.03	*
Medications
Mycophenolate Mofetil (%)	8377	81.17	%	8290	79.68	%	0.007	**	1231	79.99	%	0.28		391	81.80	%	0.78
Cyclosporine (%)	503	4.87	%	507	4.87	%	1.00		81	5.26	%	0.55		24	5.02	%	0.97
Tacrolimus (%)	9431	91.39	%	9448	90.81	%	0.15		1394	90.58	%	0.32		425	88.91	%	0.07
Sirolimus (%)	233	2.26	%	231	2.22	%	0.89		35	2.27	%	1.00		12	2.51	%	0.84
Steroids (%)	9236	89.50	%	9160	88.04	%	0.001	**	1352	87.85	%	0.06		415	86.82	%	0.08
Biomarkers
Albumin (mg/dL)	3.08	± 0.71	mg/dL	3.07	± 0.72	mg/dL	0.59		3.11	± 0.73	mg/dL	0.07		2.99	± 0.68	mg/dL	0.02	*
Creatinine (mg/dL)	1.33	± 0.96	mg/dL	1.33	± 0.96	mg/dL	0.98		1.29	± 0.94	mg/dL	0.09		1.22	± 0.79	mg/dL	0.02	*
INR (unit)	1.87	± 1.05	unit	1.87	± 1.04	unit	0.28		1.84	± 1.27	unit	< 0.001	***	1.81	± 1.01	unit	0.06
Total Bilirubin (mg/dL)	7.50	± 10.10	mg/dL	7.30	± 9.70	mg/dL	0.13		6.62	± 9.28	mg/dL	< 0.001	***	6.56	± 8.56	mg/dL	0.29
Life Support Variables
Artificial Liver Devices (%)	2	0.02	%	2	0.02	%	1.00	†	0	0.00	%	1.00	†	1	0.21	%	0.13	†
Primary Inotropic Agent							< 0.001	***				0.001	**				< 0.001	***
Dobutamine (%)	203	1.97	%	229	2.20	%			45	2.92	%			27	5.65	%
Dopamine (%)	1661	16.09	%	1483	14.25	%			213	13.84	%			74	15.48	%
Epinephrine (%)	116	1.12	%	118	1.13	%			16	1.04	%			6	1.26	%
Levophed (%)	1899	18.40	%	1822	17.51	%			305	19.82	%			78	16.32	%
Neosynephrine (%)	1674	16.22	%	1957	18.81	%			279	18.13	%			85	17.78	%
None (%)	4586	44.44	%	4556	43.79	%			641	41.65	%			189	39.54	%
Other (%)	181	1.75	%	239	2.30	%			40	2.60	%			19	3.97	%
Secondary Inotropic Agent							0.33					0.04	*				0.006	† **
Dobutamine (%)	57	0.55	%	65	0.62	%			8	0.52	%			2	0.42	%
Dopamine (%)	109	1.06	%	96	0.92	%			17	1.10	%			1	0.21	%
Epinephrine (%)	63	0.61	%	68	0.65	%			11	0.71	%			3	0.63	%
Levophed (%)	399	3.87	%	400	3.84	%			57	3.70	%			30	6.28	%
Neosynephrine (%)	470	4.55	%	535	5.14	%			102	6.63	%			33	6.90	%
None (%)	9062	87.81	%	9057	87.05	%			1321	85.83	%			399	83.47	%
Other (%)	160	1.55	%	183	1.76	%			23	1.49	%			10	2.09	%
Ternary Inotropic Agent							0.15					0.24	†				0.75	†
Dobutamine (%)	12	0.12	%	13	0.12	%			3	0.19	%			0	0.00	%
Dopamine (%)	5	0.05	%	6	0.06	%			2	0.13	%			0	0.00	%
Epinephrine (%)	22	0.21	%	17	0.16	%			1	0.06	%			0	0.00	%
Levophed (%)	47	0.46	%	47	0.45	%			10	0.65	%			3	0.63	%
Neosynephrine (%)	49	0.47	%	48	0.46	%			9	0.58	%			2	0.42	%
None (%)	10152	98.37	%	10211	98.14	%			1506	97.86	%			470	98.33	%
Other (%)	33	0.32	%	62	0.60	%			8	0.52	%			3	0.63	%
ICU Admission							0.17					0.05					0.43
ICU (%)	987	9.56	%	937	9.01	%			123	7.99	%			40	8.37	%
No ICU (%)	9333	90.44	%	9467	90.99	%			1416	92.01	%			438	91.63	%
Ventilator Support (%)	335	3.25	%	333	3.20	%	0.88		38	2.47	%	0.12		17	3.56	%	0.81
TIPS Procedure (%)	1078	10.45	%	990	9.52	%	0.03	*	141	9.16	%	0.13		55	11.51	%	0.51
Donor Demographics
Donor Age (years)	49.20	± 15.10	years	50.40	± 14.70	years	< 0.001	***	47.60	± 14.60	years	< 0.001	***	45.00	± 14.60	years	< 0.001	***
Donor Gender (%)	5605	54.31	%	5634	54.15	%	0.83		770	50.03	%	0.002	**	239	50.00	%	0.07
Donor Race							0.004	**				< 0.001	***				0.02	† *
White (%)	6855	66.42	%	7035	67.62	%			1021	66.34	%			329	68.83	%
Black (%)	1970	19.09	%	1793	17.23	%			211	13.71	%			75	15.69	%
Hispanic (%)	1048	10.16	%	1149	11.04	%			231	15.01	%			58	12.13	%
Asian (%)	256	2.48	%	250	2.40	%			30	1.95	%			14	2.93	%
Other (%)	191	1.85	%	177	1.70	%			46	2.99	%			2	0.42	%
Donor BMI (kg/m²)	29.10	± 7.32	kg/m²	30.30	± 7.50	kg/m²	< 0.001	***	31.00	± 7.44	kg/m²	< 0.001	***	31.20	± 6.83	kg/m²	< 0.001	***
Donor Biomarkers
Donor Creatinine (mg/dL)	1.91	± 1.99	mg/dL	1.86	± 1.88	mg/dL	0.13		1.91	± 1.78	mg/dL	< 0.001	***	2.01	± 1.93	mg/dL	< 0.001	***
Donor Total Bilirubin (mg/dL)	0.88	± 0.90	mg/dL	0.90	± 0.92	mg/dL	0.001	**	0.93	± 0.77	mg/dL	< 0.001	***	1.02	± 0.98	mg/dL	< 0.001	***

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†Fisher's Test

Isolative sample indicates that patients had no alternative liver diagnoses except NASH; hence, patients with hepatitis B, C, alcoholic liver disease, and HCC were excluded

There were 9184 with no macro-steatosis (control), 150 with grade 3 macro-steatosis, 822 with grade 2 macro-steatosis and 12 585 with grade 1 macro-steatosis. In terms of recipient demographics, our data showed that patients who received livers with grade 2 (73.2% vs. 67.6%, p = .001) and grade 1 (71.3% vs. 67.6%, p < .001) macro-steatosis were more likely to be male compared to the control group. Regarding co-morbidities, recipients of a liver with grade 2 macro-steatosis were less likely to have hepatitis C (38.6% vs. 44.1%, p = .003) and more likely to have alcoholic liver disease (28.2% vs. 24.2%, p = .01). Similarly, recipients of a liver with grade 1 macro-steatosis showed an increased likelihood to have alcoholic liver disease (26.8% vs. 24.2%, p < .001).

When analysing patient hepatic variables, grade 2 macro-steatosis donor liver patients had significantly lower MELD scores (18.40 ± 8.42 vs. 20.10 ± 9.62, p < .001). Grade 2 macro-steatosis donor liver patients also had lower rates of using cyclosporine (2.92% vs. 4.99%, p = .01), sirolimus (1.09% vs. 2.43%, p = .02) and steroids (85.9% vs. 89.8%, p < .001). Grade 1 macro-steatosis donor liver patients showed lower rates of using steroids (88% vs. 89.8%, p < .001) as well as lower rates of slight ascites (45.5% vs. 47.3%, p = .03) and higher rates of moderate ascites (28.2% vs. 27.4%, p = .03). In terms of biomarkers, recipients of grade 2 macro-steatosis livers had significantly lower creatinine (1.18 ± .75 mg/dL vs. 1.31 ± .96 mg/dL, p < .001), INR (1.74 ± 1.03 vs. 1.87 ± 1.04, p < .001) and total bilirubin (6.21 ± 8.19 mg/dL vs. 7.53 ± 10.10 mg/dL, p = .003). Furthermore, when analysing life support variables, patients with grade 2 macro-steatosis donor livers had a higher rate of patients which did not require inotropic agent (45.7% vs. 44.6%, p = .02), while grade 3 (34.7%, p < .001) and grade 1 (43.3%, p < .001) macro-steatosis donor livers showed lower rates compared to the control (44.6%). Additionally, patients with grade 2 macro-steatosis donor livers showed significantly fewer ICU admissions (6.81% vs. 9.44%, p = .02) and were less likely to utilize a ventilator (1.95% vs. 3.38%, p = .04).

Regarding donor characteristics, donors with grade 2 macro-steatosis were younger (45.5 ± 13.5 vs. 48.8 ± 15.8, p < .001) and had a significantly greater BMI (31.9 ± 7.97 kg/m2 vs. 28.3 ± 7.02 kg/m2, p < .001). Donors with grade 1 macro-steatosis were older (50.4 ± 14.3 years vs. 48.8 ± 15.8 years, p < .001), more likely to be male (54.7% vs. 52.7%, p = .004) and had a significantly greater BMI (30.80 ± 7.51 kg/m2 vs. 28.3 ± 7.02, p < .001). Additionally, donors with grade 2 (68%, p = <.001) and grade 1 (68.5%, p < .001) macro-steatosis had more white donors compared to the control group (64.9%).

There were 10,320 individuals without micro-steatosis (control), 478 with grade 3 micro-steatosis, 1539 with grade 2 micro-steatosis and 10 404 with grade 1 micro-steatosis. In terms of recipient demographics, patients of grade 3 micro-steatosis donor livers were younger (55.3 ± 9.16 years vs. 56.1 ± 9.45 years, p = .04) and patients of grades 2 (71.7% vs. 68.8%, p = .03) and 1 (70.6% vs. 68.8%, p = .006) micro-steatosis livers were more likely to be male. Recipients of grade 1 micro-steatosis livers were more likely to have the co-morbidity of alcoholic liver disease (26.5% vs. 25.1%, p = .002). Patients of grade 2 (19.30 ± 9.31 vs. 20.20 ± 9.57, p < .001) and grade 3 (19.2 ± 9.04 vs. 20.20 ± 9.57, p = .03) micro-steatosis livers had significantly lower MELD scores. Additionally, patients with grade 3 (34.1%, p = .002) and grade 1 (28.4%, p < .001) micro-steatosis donor livers had higher rates of moderate ascites compared to the control group (26.9%), while those with grade 2 macro-steatosis donor livers showed a higher rate of encephalopathy of 3–4% (10.5% vs. 9.84%) and lower rates of encephalopathy of 1%–2% (48.7% vs. 53.1%) or no encephalopathy (40.8% vs. 37%) compared to the control (p = .005). Regarding medications, grade 1 micro-steatosis donor liver patients had lower rates of using steroids (88% vs. 89.5%, p = .001). Furthermore, grade 3 micro-steatosis donor liver patients had lower albumin (2.99 ± .68 vs. 3.08 ± .71, p = .02) and creatinine (1.22 ± .79 mg/dL vs. 1.33 ± .96 mg/dL, p = .02) and grade 2 micro-steatosis donor liver patients had lower INR (1.84 ± 1.27 vs. 1.87 ± 1.05, p < .001) and total bilirubin (6.62 ± 9.28 mg/dL vs. 7.50 ± 10.10 mg/dL, p < .001). In regard to life support variables, grade 3 (39.5%, p < .001), grade 2 (41.7%, p = .001) and grade 1 (43.8%, p < .001) micro-steatosis donor liver all had a lower percentage of cases which did not require inotropic agents compared to the control group (44.4%).

Regarding donor variables, patients with grade 3 (45.0 ± 14.6 years vs. 49.2 ± 15.1 years, p < .001) and grade 2 (47.6 ± 14.6 years vs. 49.2 ± 15.1 years, p < .001) micro-steatosis donors were younger, while grade 1 (50.40 ± 14.7 years vs. 49.2 ± 15.1 years, p < .001) micro-steatosis donors were older. Grade 2 (50%, p = .002) and grade 1 (45.8%, p = .004) micro-steatosis donors were more likely to be female compared to the control group (45.7%). Grade 3 (31.2 ± 6.83 kg/m2, p < .001), grade 2 (31 ± 7.44 kg/m2, p < .001) and grade 1 (30.3 ± 7.5 kg/m2, p < .001) micro-steatosis donors all had a greater BMI than that of the control patients (29.1 ± 7.32 kg/m2). Grade 3 micro-steatosis donors had higher creatinine values (2.01 ± 1.93 mg/dL, p < .001), while grade 2 micro-steatosis donors had near similar creatinine values (1.91 ± 1.78 mg/dL, p < .001) when compared to the control group (1.91 ± 1.99 mg/dL). Grade 3 (1.02 ± .98 mg/dL, p < .001), grade 2 (.93 ± .77 mg/dL, p < .001) and grade 1 (.90 ± .92 mg/dL, p = .003) micro-steatosis donors all had greater total bilirubin than that of the control patients (.88 ± .90 mg/dL).

Baseline characteristics of the study cohort further stratified by the presence of mixed steatosis versus pure micro-and macro-steatosis in the donor livers can be identified in Supplementary Tables S2 and S5.

Clinical Outcomes

Table 2A analyses the impact of donor m acro-steatosis (Table 2A) and micro-steatosis (Table 2B) on all-cause mortality and graft failure. Our data show that in the final iterations of our sequential Cox regression analyses, any percentage of macro-steatosis and micro-steatosis in a transplanted liver has no significant difference on mortality or graft failure when compared to those with no steatosis. This is furthermore demonstrated in Figures 2 and 3, which show the cumulative hazard curves for the different macro-vesicular and micro-vesicular steatosis cohorts respectively. In addition, the final models of the iterative Cox regression analyses performed for all-cause mortality and graft failure are represented by the combined multivariable forest plots demonstrated in the Supplementary Figures S1 and S2 respectively.

Table 2A:

Incidence Rates and Sequential Cox Regression Analysis Using Macrosteatosis as a Prognostic Risk Factor for All-Cause Mortality and Graft Failure

Grade 1 Macrovesicular Steatosis				Grade 2 Macrovesicular Steatosis				Grade 3 Macrovesicular Steatosis
(A) All-Cause Mortality
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	3139	59.68	(57.68–61.74)	Grade 2 Macrovesicular Steatosis	162	58.42	(49.98–67.81)	Grade 3 Macrovesicular Steatosis	32	49.54	(34.13–69.22)
No Steatosis	2310	55.55	(53.36–57.79)	No Steatosis	2310	55.55	(53.36–57.79)	No Steatosis	2310	55.55	(53.36–57.79)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.04 *	1.06	(1.00–1.12)	1	0.90	0.99	(0.84–1.16)	1	0.53	0.89	(0.63–1.27)
2	0.04 *	1.06	(1.00–1.12)	2	0.96	1.00	(0.86–1.18)	2	0.49	0.88	(0.62–1.25)
3	0.04 *	1.06	(1.00–1.12)	3	0.85	1.02	(0.87–1.19)	3	0.45	0.87	(0.62–1.24)
† FM	0.18	1.04	(0.98–1.10)	† FM	0.71	1.03	(0.88–1.21)	† FM	0.50	0.89	(0.62–1.26)
(B) Graft Failure
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	690	13.12	(12.17–14.13)	Grade 2 Macrovesicular Steatosis	42	15.15	(10.94–20.42)	Grade 3 Macrovesicular Steatosis	11	17.03	(8.53–30.27)
No Steatosis	498	11.98	(10.95–13.07)	No Steatosis	498	11.98	(10.95–13.07)	No Steatosis	498	11.98	(10.95–13.07)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.26	1.07	(0.95–1.20)	1	0.40	1.15	(0.84–1.57)	1	0.31	1.36	(0.75–2.48)
2	0.22	1.07	(0.96–1.21)	2	0.30	1.18	(0.86–1.62)	2	0.33	1.34	(0.74–2.44)
3	0.22	1.08	(0.96–1.21)	3	0.25	1.20	(0.88–1.65)	3	0.34	1.34	(0.74–2.44)
† FM	0.66	1.03	(0.91–1.15)	† FM	0.23	1.22	(0.88–1.68)	† FM	0.34	1.34	(0.74–2.45)

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†FM indicates Final Model

Footnote:

^*Model 1 includes VOI (variable of interest) and demographics; Model 2 includes Model 1 terms with the addition of comorbidities, and liver disease etiologies; Model 3 includes Model 2 terms with the addition of hepatic variables, MELD score, and liver laboratory markers; Model 4 includes Model 3 terms with the addition of donor demographics

Table 2B:

Incidence Rates and Sequential Cox Regression Analysis Using Microsteatosis as a Prognostic Risk Factor for All-Cause Mortality and Graft Failure

Grade 1 Microvesicular Steatosis				Grade 2 Microvesicular Steatosis				Grade 3 Microvesicular Steatosis
(A) All-Cause Mortality
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	2566	59.38	(57.17–61.64)	Grade 2 Microvesicular Steatosis	335	57.05	(51.26–63.30)	Grade 3 Microvesicular Steatosis	111	53.41	(44.14–63.96)
No Steatosis	2631	56.66	(54.58–58.80)	No Steatosis	2631	56.66	(54.58–58.80)	No Steatosis	2631	56.66	(54.58–58.80)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.32	1.03	(0.97–1.09)	1	0.56	0.97	(0.86–1.08)	1	0.52	0.94	(0.78–1.14)
2	0.33	1.03	(0.97–1.09)	2	0.63	0.97	(0.87–1.09)	2	0.49	0.94	(0.77–1.13)
3	0.33	1.03	(0.97–1.08)	3	0.73	0.98	(0.87–1.10)	3	0.47	0.93	(0.77–1.13)
† FM	0.65	1.01	(0.96–1.07)	† FM	0.83	0.99	(0.88–1.11)	† FM	0.86	0.98	(0.81–1.19)
(B) Graft Failure
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	576	13.33	(12.27–14.45)	Grade 2 Microvesicular Steatosis	80	13.63	(10.82–16.93)	Grade 3 Microvesicular Steatosis	20	9.62	(5.89–14.82)
No Steatosis	565	12.17	(11.19–13.21)	No Steatosis	565	12.17	(11.19–13.21)	No Steatosis	565	12.17	(11.19–13.21)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.34	1.06	(0.94–1.19)	1	0.85	1.02	(0.81–1.29)	1	0.20	0.75	(0.48–1.17)
2	0.33	1.06	(0.94–1.19)	2	0.81	1.03	(0.81–1.30)	2	0.18	0.74	(0.47–1.15)
3	0.34	1.06	(0.94–1.19)	3	0.71	1.05	(0.83–1.32)	3	0.20	0.75	(0.48–1.17)
† FM	0.66	1.03	(0.91–1.15)	† FM	0.69	1.05	(0.83–1.33)	† FM	0.35	0.81	(0.51–1.26)

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†FM indicates Final Model

Footnote:

^*Model 1 includes VOI (variable of interest) and demographics; Model 2 includes Model 1 terms with the addition of comorbidities, and liver disease etiologies; Model 3 includes Model 2 terms with the addition of hepatic variables, MELD score, and liver laboratory markers; Model 4 includes Model 3 terms with the addition of donor demographics

Figure 2. Comparison of Cumulative Events with All-Cause Mortality and Graft Failure as Endpoint Macrovesicular Steatosis.

This figure shows the differences in cumulative hazards using all-cause mortality and graft failure as primary endpoints, in liver transplant recipients who received grafts from donors with varying degrees of hepatic macrovesicular steatosis (A,B), mixed-steatosis (C,D), and pure macrosteatosis (E,F).

(A) and (B) represent the cumulative hazard(s) for all-cause mortality and graft failure, respectively, in liver transplant recipients, whoreceived their grafts from donors with grade 1 macrovesicular steatosis versus donors with grade 2 macrovesicular steatosis versus donorswith grade 3 macrovesicular steatosis versus donors with no steatosis. (C) and (D) represent the cumulative hazard(s) for all-cause mortalityand graft failure, respectively, in liver transplant recipients, who received their grafts from donors with grade 1 mixed-macrovesicular steatosisversus donors with grade 2 mixed-macrovesicular steatosis versus grade 3 mixed-macrovesicular steatosis. (E) and (F) represent thecumulative hazard(s) for all-cause mortality and graft failure, respectively, in liver transplant recipients, who received their grafts from donorswith grade 1 pure macrovesicular steatosis versus donors with grade 2 pure macrovesicular steatosis versus grade 3 pure macrovesicularsteatosis versus donors with no steatosis.

Figure 3. Comparison of Cumulative Events with All-Cause Mortality and Graft Failure as Endpoint Microvesicular Steatosis.

This figure shows the differences in cumulative hazards using all-cause mortality and graft failure as primary endpoints, in liver transplant recipients who received grafts from donors with varying degrees of hepatic microvesicular steatosis (A,B), mixed-steatosis (C,D), and pure microsteatosis (E,F).

(A) and (B) represent the cumulative hazard(s) for all-cause mortality and graft failure, respectively, in liver transplant recipients, whoreceived their grafts from donors with grade 1 microvesicular steatosis versus donors with grade 2 microvesicular steatosis versus donorswith grade 3 microvesicular steatosis versus donors with no steatosis. (C) and (D) represent the cumulative hazard(s) for all-cause mortalityand graft failure, respectively, in liver transplant recipients, who received their grafts from donors with grade 1 mixed-microvesicular steatosisversus donors with grade 2 mixed-microvesicular steatosis versus grade 3 mixed-microvesicular steatosis. (E) and (F) represent thecumulative hazard(s) for all-cause mortality and graft failure, respectively, in liver transplant recipients, who received their grafts from donorswith grade 1 pure microvesicular steatosis versus donors with grade 2 pure microvesicular steatosis versus grade 3 pure microvesicularsteatosis versus donors with no steatosis.

However, Table 3A reveals that macro-steatosis (Table 3A) and micro-steatosis (Table 3B) do have significantly increased rates of mortality when stratified by specific causes of death. Transplant recipients of grade 2 macro-steatosis livers had a statistically greater incidence of cardiac arrest (aHR 1.62, p = .03) as a cause of death. Patients who received a donor organ with grade 1 macro-steatosis had a statistically greater incidence of mortality due to graft rejection (aHR 1.78, p = .006). Regarding micro-steatosis, recipients of grade 1 micro-steatosis donor livers had greater mortality due to graft rejection (aHR 1.61, p = .02). Supplementary Table S1 includes results from our competing risk regression analysis using macro-steatosis and micro-steatosis as prognostic risk factors for all-cause mortality and graft failure.

Table 3A:

Incidence Rates and Sequential Cox Regression Analysis Using Macrosteatosis as a Prognostic Risk Factor for Specific Causes of Death

Grade 1 Macrovesicular Steatosis				Grade 2 Macrovesicular Steatosis				Grade 3 Macrovesicular Steatosis
(A) Death due to Cardiac Arrest
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	264	5.02	(4.43–5.66)	Grade 2 Macrovesicular Steatosis	23	8.29	(5.27-12.42)	Grade 3 Macrovesicular Steatosis	4	6.19	(1.69–15.78)
No Steatosis	176	4.23	(3.63–4.90)	No Steatosis	176	4.23	(3.63–4.90)	No Steatosis	176	4.23	(3.63–4.90)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.18	1.14	(0.94–1.38)	1	0.02 *	1.71	(1.11–2.64)	1	0.47	1.44	(0.53–3.89)
2	0.20	1.13	(0.94–1.37)	2	0.02 *	1.69	(1.09–2.62)	2	0.51	1.40	(0.52–3.77)
3	0.21	1.13	(0.93–1.37)	3	0.01 *	1.75	(1.13–2.71)	3	0.54	1.36	(0.50–3.68)
† FM	0.42	1.08	(0.89–1.32)	† FM	0.03 *	1.62	(1.04–2.55)	† FM	0.55	1.36	(0.50–3.67)
(B) Death due to General Cardiac Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	425	8.08	(7.33–8.88)	Grade 2 Macrovesicular Steatosis	31	11.18	(7.61–15.83)	Grade 3 Macrovesicular Steatosis	6	9.29	(3.42–20.11)
No Steatosis	294	7.07	(6.29–7.92)	No Steatosis	294	7.07	(6.29–7.92)	No Steatosis	294	7.07	(6.29–7.92)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.20	1.10	(0.95–1.28)	1	0.07	1.40	(0.97–2.03)	1	0.50	1.32	(0.59–2.97)
2	0.23	1.10	(0.94–1.27)	2	0.08	1.39	(0.96–2.02)	2	0.55	1.28	(0.57–2.87)
3	0.24	1.09	(0.94–1.27)	3	0.05	1.44	(1.00–2.10)	3	0.56	1.27	(0.57–2.86)
† FM	0.35	1.07	(0.92–1.25)	† FM	0.08	1.41	(0.96–2.06)	† FM	0.56	1.27	(0.57–2.87)
(D) Death due to Graft Rejection
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	76	1.45	(1.14–1.81)	Grade 2 Macrovesicular Steatosis	5	1.80	(0.59–4.20)	Grade 3 Macrovesicular Steatosis	0	0.00	(0.00–5.69)
No Steatosis	34	0.82	(0.57–1.14)	No Steatosis	34	0.82	(0.57–1.14)	No Steatosis	34	0.82	(0.57–1.14)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.006 **	1.77	(1.18–2.65)	1	0.14	2.02	(0.79–5.18)	1	1.00	0.00	(0.00-Inf)
2	0.006 **	1.77	(1.18–2.65)	2	0.14	2.03	(0.79–5.20)	2	1.00	0.00	(0.00-Inf)
3	0.006 **	1.76	(1.17–2.63)	3	0.11	2.18	(0.84–5.63)	3	1.00	0.00	(0.00-Inf)
† FM	0.006 **	1.78	(1.18–2.68)	† FM	0.18	1.94	(0.73–5.18)	† FM	1.00	0.00	(0.00-Inf)
(F) Death due to General Infectious Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	545	10.36	(9.51–11.27)	Grade 2 Macrovesicular Steatosis	24	8.66	(5.55–12.85)	Grade 3 Macrovesicular Steatosis	3	4.64	(0.96–13.51)
No Steatosis	409	9.83	(8.91–10.83)	No Steatosis	409	9.83	(8.91–10.83)	No Steatosis	409	9.83	(8.91–10.83)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.69	1.03	(0.90–1.17)	1	0.24	0.78	(0.52–1.18)	1	0.17	0.45	(0.15–1.41)
2	0.71	1.02	(0.90–1.17)	2	0.24	0.78	(0.52–1.18)	2	0.17	0.45	(0.14–1.40)
3	0.69	1.03	(0.90–1.17)	3	0.34	0.82	(0.54–1.24)	3	0.15	0.44	(0.14–1.37)
† FM	0.92	1.01	(0.88–1.15)	† FM	0.61	0.90	(0.59–1.36)	† FM	0.19	0.47	(0.15–1.46)
(G) Death due to General Neurological Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	95	1.81	(1.46–2.21)	Grade 2 Macrovesicular Steatosis	2	0.72	(0.09–2.60)	Grade 3 Macrovesicular Steatosis	1	1.55	(0.04–8.60)
No Steatosis	78	1.88	(1.48–2.34)	No Steatosis	78	1.88	(1.48–2.34)	No Steatosis	78	1.88	(1.48–2.34)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.75	0.95	(0.71–1.28)	1	0.15	0.36	(0.09–1.46)	1	0.88	0.86	(0.12–6.18)
2	0.74	0.95	(0.70–1.28)	2	0.15	0.36	(0.09–1.46)	2	0.88	0.86	(0.12–6.19)
3	0.74	0.95	(0.70–1.28)	3	0.17	0.37	(0.09–1.53)	3	0.87	0.85	(0.12–6.14)
† FM	0.61	0.92	(0.68–1.25)	† FM	0.25	0.44	(0.11–1.80)	† FM	0.88	0.86	(0.12–6.22)
(H) Death due to General Renal Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	143	2.72	(2.29–3.20)	Grade 2 Macrovesicular Steatosis	9	3.25	(1.49–6.15)	Grade 3 Macrovesicular Steatosis	2	3.10	(0.38–11.14)
No Steatosis	124	2.98	(2.48–3.55)	No Steatosis	124	2.98	(2.48–3.55)	No Steatosis	124	2.98	(2.48–3.55)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.34	0.89	(0.70–1.13)	1	0.99	1.01	(0.51–1.98)	1	0.97	1.03	(0.25–4.15)
2	0.35	0.89	(0.70–1.14)	2	0.90	1.04	(0.53–2.06)	2	1.00	1.00	(0.25–4.06)
3	0.35	0.89	(0.70–1.13)	3	0.78	1.10	(0.56–2.18)	3	0.94	0.95	(0.23–3.86)
† FM	0.28	0.87	(0.68–1.12)	† FM	0.82	1.09	(0.54–2.17)	† FM	0.95	0.96	(0.24–3.89)
(I) Death due to General Respiratory Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	197	3.75	(3.24–4.31)	Grade 2 Macrovesicular Steatosis	12	4.33	(2.24–7.55)	Grade 3 Macrovesicular Steatosis	2	3.10	(0.38–11.14)
No Steatosis	146	3.51	(2.97–4.13)	No Steatosis	146	3.51	(2.97–4.13)	No Steatosis	146	3.51	(2.97–4.13)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.66	1.05	(0.85–1.30)	1	0.73	1.11	(0.61–2.00)	1	0.85	0.87	(0.22–3.52)
2	0.69	1.05	(0.84–1.30)	2	0.72	1.11	(0.62–2.01)	2	0.84	0.86	(0.21–3.49)
3	0.65	1.05	(0.85–1.30)	3	0.66	1.14	(0.63–2.06)	3	0.80	0.84	(0.21–3.38)
† FM	0.54	1.07	(0.86–1.33)	† FM	0.90	1.04	(0.57–1.90)	† FM	0.78	0.82	(0.20–3.33)
(J) Death due to Sepsis
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Macrovesicular Steatosis	348	6.62	(5.94–7.35)	Grade 2 Macrovesicular Steatosis	16	5.77	(3.30–9.35)	Grade 3 Macrovesicular Steatosis	3	4.64	(0.96–13.51)
No Steatosis	257	6.18	(5.45–6.98)	No Steatosis	257	6.18	(5.45–6.98)	No Steatosis	257	6.18	(5.45–6.98)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.66	1.04	(0.88–1.22)	1	0.43	0.81	(0.49–1.35)	1	0.59	0.73	(0.23–2.28)
2	0.67	1.04	(0.88–1.22)	2	0.44	0.82	(0.50–1.36)	2	0.59	0.73	(0.23–2.28)
3	0.65	1.04	(0.88–1.22)	3	0.54	0.85	(0.51–1.41)	3	0.56	0.71	(0.23–2.23)
† FM	0.80	1.02	(0.87–1.20)	† FM	0.93	0.98	(0.58–1.63)	† FM	0.67	0.78	(0.25–2.44)

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†FM indicates Final Model

Footnote:

^*Model 1 includes VOI (variable of interest) and demographics; Model 2 includes Model 1 terms with the addition of comorbidities, and liver disease etiologies; Model 3 includes Model 2 terms with the addition of hepatic variables, MELD score, and liver laboratory markers; Model 4 includes Model 3 terms with the addition of donor demographics

Table 3B:

Incidence Rates and Sequential Cox Regression Analysis Using Microsteatosis as a Prognostic Risk Factor for Specific Causes of Death

Grade 1 Microvesicular Steatosis				Grade 2 Microvesicular Steatosis				Grade 3 Microvesicular Steatosis
(A) Death due to Cardiac Arrest
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	225	5.21	(4.55–5.93)	Grade 2 Microvesicular Steatosis	31	5.28	(3.59–7.49)	Grade 3 Microvesicular Steatosis	7	3.37	(1.36–6.93)
No Steatosis	204	4.39	(3.81–5.04)	No Steatosis	204	4.39	(3.81–5.04)	No Steatosis	204	4.39	(3.81–5.04)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.19	1.13	(0.94–1.37)	1	0.62	1.10	(0.75–1.61)	1	0.46	0.75	(0.35–1.60)
2	0.21	1.13	(0.94–1.37)	2	0.59	1.11	(0.76–1.62)	2	0.45	0.75	(0.35–1.59)
3	0.22	1.13	(0.93–1.36)	3	0.50	1.14	(0.78–1.67)	3	0.44	0.74	(0.35–1.58)
† FM	0.34	1.10	(0.91–1.33)	† FM	0.73	1.07	(0.73–1.57)	† FM	0.37	0.71	(0.33–1.51)
(B) Death due to General Cardiac Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	349	8.08	(7.25–8.96)	Grade 2 Microvesicular Steatosis	43	7.32	(5.30–9.85)	Grade 3 Microvesicular Steatosis	16	7.70	(4.41–12.47)
No Steatosis	348	7.49	(6.73–8.32)	No Steatosis	348	7.49	(6.73–8.32)	No Steatosis	348	7.49	(6.73–8.32)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.64	1.04	(0.89–1.20)	1	0.52	0.90	(0.66–1.24)	1	0.95	1.02	(0.62–1.68)
2	0.66	1.03	(0.89–1.20)	2	0.57	0.91	(0.66–1.25)	2	0.95	1.01	(0.61–1.68)
3	0.68	1.03	(0.89–1.20)	3	0.68	0.94	(0.68–1.29)	3	0.97	1.01	(0.61–1.67)
† FM	0.78	1.02	(0.88–1.19)	† FM	0.60	0.92	(0.67–1.26)	† FM	0.96	0.99	(0.59–1.63)
(D) Death due to Graft Rejection
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	64	1.48	(1.14–1.89)	Grade 2 Microvesicular Steatosis	9	1.53	(0.70–2.91)	Grade 3 Microvesicular Steatosis	1	0.48	(0.01–2.68)
No Steatosis	41	0.88	(0.63–1.20)	No Steatosis	41	0.88	(0.63–1.20)	No Steatosis	41	0.88	(0.63–1.20)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.01 *	1.64	(1.11–2.42)	1	0.21	1.59	(0.77–3.27)	1	0.50	0.50	(0.07–3.67)
2	0.01 *	1.64	(1.10–2.42)	2	0.21	1.59	(0.77–3.27)	2	0.49	0.49	(0.07–3.60)
3	0.02 *	1.63	(1.10–2.41)	3	0.20	1.60	(0.78–3.30)	3	0.43	0.45	(0.06–3.27)
† FM	0.02 *	1.61	(1.08–2.38)	† FM	0.28	1.50	(0.72–3.12)	† FM	0.42	0.44	(0.06–3.22)
(F) Death due to General Infectious Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	473	10.95	(9.99–11.97)	Grade 2 Microvesicular Steatosis	56	9.54	(7.21–12.37)	Grade 3 Microvesicular Steatosis	11	5.29	(2.64–9.45)
No Steatosis	441	9.50	(8.64–10.42)	No Steatosis	441	9.50	(8.64–10.42)	No Steatosis	441	9.50	(8.64–10.42)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.11	1.11	(0.98–1.27)	1	0.51	0.91	(0.69–1.20)	1	0.04 *	0.54	(0.30–0.98)
2	0.11	1.11	(0.98–1.26)	2	0.52	0.91	(0.69–1.21)	2	0.04 *	0.54	(0.30–0.99)
3	0.11	1.11	(0.98–1.26)	3	0.69	0.94	(0.71–1.25)	3	0.05	0.55	(0.30–1.00)
† FM	0.18	1.09	(0.96–1.25)	† FM	0.88	0.98	(0.74–1.30)	† FM	0.12	0.62	(0.34–1.13)
(G) Death due to General Neurological Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	88	2.04	(1.63–2.51)	Grade 2 Microvesicular Steatosis	6	1.02	(0.38–2.22)	Grade 3 Microvesicular Steatosis	5	2.41	(0.78–5.61)
No Steatosis	77	1.66	(1.31–2.07)	No Steatosis	77	1.66	(1.31–2.07)	No Steatosis	77	1.66	(1.31–2.07)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.24	1.20	(0.88–1.63)	1	0.23	0.60	(0.26–1.38)	1	0.34	1.56	(0.63–3.86)
2	0.23	1.20	(0.89–1.64)	2	0.21	0.59	(0.26–1.36)	2	0.32	1.59	(0.64–3.93)
3	0.22	1.21	(0.89–1.65)	3	0.26	0.62	(0.27–1.43)	3	0.35	1.54	(0.62–3.82)
† FM	0.30	1.18	(0.86–1.60)	† FM	0.32	0.65	(0.28–1.51)	† FM	0.30	1.62	(0.65–4.05)
(H) Death due to General Renal Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	124	2.87	(2.39–3.42)	Grade 2 Microvesicular Steatosis	12	2.04	(1.06–3.57)	Grade 3 Microvesicular Steatosis	4	1.92	(0.52–4.92)
No Steatosis	138	2.97	(2.50–3.51)	No Steatosis	138	2.97	(2.50–3.51)	No Steatosis	138	2.97	(2.50–3.51)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.62	0.94	(0.74–1.20)	1	0.15	0.65	(0.36–1.17)	1	0.36	0.63	(0.23–1.70)
2	0.61	0.94	(0.74–1.20)	2	0.17	0.66	(0.36–1.19)	2	0.37	0.63	(0.23–1.71)
3	0.61	0.94	(0.74–1.20)	3	0.19	0.68	(0.37–1.22)	3	0.39	0.65	(0.24–1.76)
† FM	0.57	0.93	(0.73–1.19)	† FM	0.23	0.69	(0.38–1.26)	† FM	0.44	0.68	(0.25–1.84)
(I) Death due to General Respiratory Causes
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	167	3.86	(3.30–4.50)	Grade 2 Microvesicular Steatosis	22	3.75	(2.35–5.67)	Grade 3 Microvesicular Steatosis	5	2.41	(0.78–5.61)
No Steatosis	163	3.51	(2.99–4.09)	No Steatosis	163	3.51	(2.99–4.09)	No Steatosis	163	3.51	(2.99–4.09)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.52	1.07	(0.87–1.33)	1	0.97	1.01	(0.64–1.57)	1	0.42	0.69	(0.29–1.69)
2	0.53	1.07	(0.86–1.33)	2	0.94	1.02	(0.65–1.59)	2	0.41	0.69	(0.28–1.68)
3	0.49	1.08	(0.87–1.34)	3	0.84	1.05	(0.67–1.63)	3	0.44	0.70	(0.29–1.72)
† FM	0.47	1.08	(0.87–1.35)	† FM	0.93	1.02	(0.65–1.60)	† FM	0.41	0.69	(0.28–1.69)
(J) Death due to Sepsis
	Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years			Raw Incidence	Incidence Rates per 1000 Person-Years
Grade 1 Microvesicular Steatosis	294	6.80	(6.05–7.62)	Grade 2 Microvesicular Steatosis	37	6.30	(4.44–8.68)	Grade 3 Microvesicular Steatosis	5	2.41	(0.78–5.61)
No Steatosis	288	6.20	(5.51–6.96)	No Steatosis	288	6.20	(5.51–6.96)	No Steatosis	288	6.20	(5.51–6.96)
Sequential Cox Regression Analysis
Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI	Model	p-value	aHR	95% CI
1	0.56	1.05	(0.89–1.23)	1	0.59	0.91	(0.65–1.28)	1	0.03 *	0.37	(0.15–0.90)
2	0.56	1.05	(0.89–1.23)	2	0.58	0.91	(0.64–1.28)	2	0.03 *	0.37	(0.15–0.90)
3	0.54	1.05	(0.89–1.24)	3	0.73	0.94	(0.67–1.33)	3	0.03 *	0.38	(0.16–0.92)
† FM	0.66	1.04	(0.88–1.22)	† FM	0.95	0.99	(0.70–1.40)	† FM	0.07	0.44	(0.18–1.06)

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

^†FM indicates Final Model

Footnote:

^*Model 1 includes VOI (variable of interest) and demographics; Model 2 includes Model 1 terms with the addition of comorbidities, and liver disease etiologies; Model 3 includes Model 2 terms with the addition of hepatic variables, MELD score, and liver laboratory markers; Model 4 includes Model 3 terms with the addition of donor demographics

To supplement the earlier findings, further analysis was conducted to identify the role of the presence of mixed steatosis in donor liver grafts as a prognostic risk factor for all-cause mortality, graft failure and other specific aetiologies of death. Supplementary Tables S3 and S6 demonstrate the incidence rates of the outcomes of all-cause mortality and graft failure in the patient cohorts with mixed steatosis and pure macro-and micro-steatosis respectively. Supplementary Tables S4 and S7 indicate the results from the sequential Cox regression analysis using mixed steatosis and pure macro-and micro-steatosis as prognostic risk factors for specific aetiologies of death respectively.

Discussion

This analysis evaluates the effect of various grades of hepatic macro-or micro-steatosis of a LT donor organ on post-LT outcomes, particularly highlighting notable causes of mortality which appear to influence post-LT prognosis. To our knowledge, this is the largest sample size of patients analysing hepatic steatosis in relation to LT outcomes. Additionally, this study is novel in that it includes stratification of steatosis across severity of both macro-and micro-steatosis and further specification according to various causes of mortality. This allows for identification of particularly vulnerable LT recipients to better define parameters for recipients of steatotic grafts. Additionally, the highlighted causes of mortality in our analysis can also be used to help guide management post-LT.

Our analysis found that among recipients who received a donor organ with any evidence of macro-or micro-steatosis, there was no statistically significant change in all-cause mortality or graft failure, when compared to those with no steatosis. However, analysis of specific causes of mortality revealed that among recipients that received a donor organ with grade 2 macro-steatosis, there was statistically increased mortality due to cardiac arrest. Recipients of a donor organ with grade 1 macro-steatosis had a statistically greater incidence of mortality due to transplant rejection. Regarding micro-steatosis, recipients of a donor liver with grade 1 micro-steatosis donor livers had greater mortality due to graft rejection.

Current guidance regarding the use of grafts with steatosis is unclear and relies on physician discretion. There are various permutations in the literature in regards to the quantitative degree of and pure versus mixed composition of steatosis. This study's findings of an absence of statistical difference in all-cause mortality related to the presence or severity of neither macro-or micro-steatosis as compared to donor organs without steatosis may indicate a viable donor organ option that overall does not impact post-transplant mortality regardless of the degree or pure versus mixed nature of steatosis, which is novel to this study. This finding is supported by Han et al. which demonstrates that when specifically controlling for pure micro-steatosis (5%–50%), there was no significant impact on post-operative recipient and graft survival, similar to findings in this study (17). These data are however complicated specifically in the setting of donation after circulatory death as Bath et al. reports that recipients demonstrated worse outcomes in the setting of macro-and micro-steatosis, revealing sub-groups that conflict with these conclusions from Han et al. and this study (17,18). As such when paired with the appropriate recipient with minimal co-morbidities and consideration for circulatory compromise, suboptimal grafts with evidence of steatosis may be successful with benefits outweighing the risks (19). This may further suggest that management of complications and co-morbidities play a crucial role in post-LT prognosis. Careful identification and management of vulnerabilities and patient co-morbidities may therefore allow for expansion of the pool of viable donor livers to meet the increasing demand for organs (20). Han et al. further shows that varying degrees of micro-steatosis do not impact outcomes in the mixed setting of macro-steatosis, which similar to this study demonstrates how outcomes were not influenced by the quantitative degree of steatosis (17). However, this conclusion does not rule out complications due to macro-steatosis, as demonstrated by a retrospective study by Spitzer et al. which identified macro-steatosis to be an independent risk factor for graft failure at one year.21 More specifically in the setting of ischaemia and reperfusion with intermittent hepatic inflow occlusion, significant changes in transaminases were observed in analysis of macro-steatosis, not replicated in grafts with pure micro-steatosis (22). Although transaminases do not clinically equate to all-cause mortality in this study, this raises concern regarding macro-steatosis notably is not replicated in this study. Some key differences may explain the difference in results. Spitzer et al. used a follow-up period of 1 year which is more concerning for short-term risk factors of graft failure, while our analysis evaluates risk in more long-term LT prognosis with median follow-up times of 3.53–3.70 years (21). Additionally, graft failure in their study was defined as either re-transplantation or recipient death, which differs from our endpoint of graft failure as registered in the UNOS database (21). As such, it is possible that macro-steatosis may still contribute to graft failure in the short term which is captured by our study, while also not affecting long-term mortality outcomes.

Our results identified mortality particularly in the setting of cardiac arrest and graft rejection. This is similar to observations in previous studies that identify cardiovascular events as a leading cause of non-graft-related mortality in LT with risks of cardiovascular events after 1 and 3 years post-LT of 4.5% and 10.1% (23). One study in 2012 by Albeldawi et al. assessed cardiovascular events among recipients with non-alcoholic steatohepatitis (NASH) as the indication for transplant and reported 15.3% and 19.3% cumulative risk of cardiovascular events (at 1 and 3 years post-LT respectively), significantly elevated from the aforementioned risk generally across all LT. Albeldawi's study also identified hypertension and diabetes, independent from NASH diagnosis, to be associated with nearly twice the risk of cardiovascular events (24). As such, it can be extrapolated that the increased oxidative stress and lipid peroxidation associated with metabolic syndromes such as NASH poses a cardiovascular risk (25). However, distinct from our analysis, these studies reveal the impact of metabolic syndrome on the hepatic graft post-LT. Our analysis shows a similar result of cardiovascular vulnerability in the setting of steatosis of the donor hepatic graft on post-LT outcomes, which may suggest persistence of these metabolic dysfunction-related risks presenting post-transplantation. As such, similar management strategies of modifiable risk factors such as diabetes, obesity and hypertension are warranted in this population of LT recipients regardless of their metabolic health.

The absence of significant difference in mortality secondary to coagulopathy in the setting of grade 3 steatosis may be the result of a more cautionary approach to the use of grafts with grade 3 macro-or micro-steatosis, especially given that literature suggests that grafts with higher grades of steatosis led more frequently to graft dysfunction. One study in particular attributed >30% macro-steatosis to be associated with a relative risk of 1.71 and as such transplant centres likely avoid use of grafts with grade 3 severity unless necessary with the benefits of a suboptimal graft outweighing its risks (21,26). If we revisit how our study revealed no statistical change in all-cause mortality due to grade 3 macro-or micro-steatotic grafts, it could be that current avoidance of higher grade steatotic grafts optimally removes the grafts most likely to fail from our study and thus optimal management of grade 3 steatosis requires further investigation.

Our findings of specific causes of death according to type of steatosis and severity reveals what specific qualities of a recipient should be evaluated to mitigate complications and mortality, as well as to identify optimal recipients for these suboptimal grafts. Limitations of this study include the retrospective model of study, which is subject to selection bias and limited collected data to expose confounding factors related to mortality and definitions may vary between entries. As such, the UNOS-STAR database may not capture all details or confounders that influence mortality seen in this population. In order to minimize bias, cohorts were matched and adjusted for donor demographics, co-morbidities and donor and recipient profiles. The modality through which severity of hepatic steatosis was determined poses another possible limitation. Data regarding the presence of micro-steatosis or macro-steatosis were obtained through the UNOS registry. UNOS requires institutions to stain pathology slides through standardized protocols to discern severity of steatosis. However, there may have been variable interpretation techniques among each participating institution and pathologist, thereby introducing a source of bias. Furthermore, biopsy data regarding presence and severity of fibrosis were unfortunately not available in the registry, findings of which may have a notable impact on graft outcomes. Although histological presence of fibrosis was not available, the prevalence of cirrhosis and diagnostic indications for orthotopic liver transplantation were included in the univariate analysis. Finally, a limitation in interpretation of these results is the lack of data to correlate mortality to time post-LT. Given that more likely rejection types and immune responses differ according to the amount of time post-LT, it would be important to investigate how causes of mortality may or may not be affected during various stages or amount of time post-LT in future studies.

The need for hepatic grafts is growing and the proportion of the population impacted by hepatic steatosis is expanding; it is important to consider the use of steatotic hepatic grafts in the setting of limited optimal donor organs, especially given the high risk of mortality from end-stage hepatic dysfunction (20,27). This study importantly identifies prognostic and mortality factors that pose a risk to LT recipients of steatotic grafts. With results which show no significant change in mortality but rather significance in causes of mortality including cardiac arrest and graft rejection management can be directed towards preventing and identifying such risks as it pertains to management in the setting of a graft with hepatic steatosis.

Conclusion

This study attempted to identify the effect of the presence and severity of hepatic steatosis of a donor liver on post-transplant outcomes. Our analysis demonstrated no significant difference in all-cause mortality or graft failure among recipients who received a donor liver with any degree of micro-or macro-steatosis. Further analysis identified significantly increased mortality due to specific aetiologies, including cardiac and graft rejection among recipients receiving donor organs with varying grades of macro-and micro-steatosis.

Supplementary Material

Supplementary Figure 1

Supplementary Figure 1. Forest plot of adjusted hazard ratios from the final model of the Cox regression analysis graft failure as the outcome in a cohort of liver transplant recipients who received grafts with donor macrovesicular steatosis.

Supplementary Figure 2

Supplementary Figure 2. Forest plot of adjusted hazard ratios from the final model of the Cox regression analysis graft failure as the outcome in a cohort of liver transplant recipients who received grafts with donor microvesicular steatosis.