Survival Benefit of Accepting Livers from Deceased Donors over 70 Years Old

Abstract

Livers from older donors (OLD; age≥70) are risky and often declined; however, it is likely that some candidates will benefit from OLDs versus waiting for younger ones. To characterize the survival benefit of accepting OLD grafts, we used 2009–2017 SRTR data to identify 24,431 adult liver transplant (LT) candidates who were offered OLD grafts eventually accepted by someone. Outcomes from the time-of-offer were compared between candidates who accepted an OLD graft and matched controls within MELD±2 who declined the same offer. Candidates who accepted OLD grafts (n=1311) were older (60.5 vs. 57.8 years, p<0.001), had a higher median MELD score (25 vs. 22, p<0.001), and were less likely to have hepatitis C cirrhosis (14.9% vs 31.2%, p<0.001). Five-year cumulative mortality among those who accepted vs declined the same OLD offer was 23.4% vs 41.2% (p<0.001). Candidates who accepted OLDs experienced an almost 2-fold reduction in mortality (aHR:_0.450.52_0.59,p<0.001) compared to those who declined the same offer, especially among the highest MELD (35–40) candidates (aHR: _0.100.24_0.55,p=0.001). Accepting an OLD offer provided substantial long-term survival benefit compared to waiting for a better organ offer, notably among candidates with MELD 35–40. Providers should consider these benefits as they evaluate OLD graft offers.

INTRODUCTION

Over 14,000 patients are currently on the U.S. liver transplant (LT) waitlist, yet only 8,082 LTs were performed in 2017 (1). More than 10% of liver waitlist candidates die awaiting transplantation (2), but 84% of candidates who died had declined at least one liver offer (3). It is likely many of these declined offers were marginal livers from older donors, considered too risky for the candidate. Older liver donors (OLD) grafts remain a critically underutilized resource (4, 5), with utilization variation between regions ranging from 0.9–12.9% (5), likely due to uncertainty about the survival benefit of these organs. However, increased OLD graft use may be a potential way to expand the donor pool, mitigate waitlist mortality, and improve survival for waitlist candidates.

Historically, the evaluation of an older donor liver has been made in the context of increased risk of mortality and graft loss associated with these organs (6–10) and decreased survival benefit of acceptance with increasing donor age (11). Yet patients and their providers are rarely faced with simultaneous offers of an older and a younger donor, in which case the decision would be clear. Instead, they are faced with the decision to accept an older donor graft today or to decline the graft and remain on the waitlist waiting for a younger donor offer that might never come. We hypothesized that, while OLD grafts carry increased risk of poor post-transplant outcomes (6–10), the benefit of the transplant may outweigh these risks; quantification of the survival benefit for OLD grafts is critical, particularly in light of decreased utilization and recovery of OLD grafts over time (12).

To better inform clinical decision-making regarding OLD grafts, we sought to characterize OLD acceptance practices and the impact of these decisions on patient mortality. We used national registry data to 1) study candidates who declined an OLD graft, 2) characterize clinical outcomes after an OLD graft decline, 3) quantify changes in disease severity (as measured by MELD score) after OLD graft decline, and 4) estimate survival benefit associated with OLD graft offer acceptance.

METHODS

Data Source

This study used data from the Scientific Registry of Transplant Recipients (SRTR) external release made available in September 2018. The SRTR data system includes data on all donors, waitlisted candidates, and transplant recipients in the US submitted by members of the Organ Procurement and Transplantation Network (OPTN), and has been described elsewhere (13). The Health and Resources Services Administration, US Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors.

Study population

We identified 24,431 non-Status 1 adult liver-only transplant candidates with MELD ≥15 who were offered an OLD graft (donor age ≥70) that was eventually used for transplantation between 12/28/2009 to 1/6/2017. Of the 24,431 candidates, 1,311 eventually accepted and were transplanted with an OLD graft (23,709 declined). OLD graft decline is subsequently referred to as candidate decline while the decision might have been made at the physician or center level. Among the 23,709 candidates who ever declined an OLD graft offer during the study period, 589 (2.4%) were later offered and accepted a different OLD graft during the study period (these candidates are included in both columns of Table 1). Candidates who listed at age<18 (n=689), were ever listed for a kidney transplant (n=1241), were inactive at the time of offer (n=157), or had died prior to the offer (n=25) were excluded. We used χ² tests and Wilcoxon rank-sum test to compare age, sex, race, MELD at offer, and indication for LT between candidates who accepted versus declined the OLD offer.

Table 1. Characteristics of candidates who accepted and declined OLD liver offers between 12/28/2009–1/6/2017.

Candidates not yet matched on MELD score.

Characteristic	Declined Offer (N=23709)	Accepted Offer (N=1311)	p value
Age at offer, median (IQR)	57.8 (51.3, 63.0)	60.5 (55.0, 65.4)	<0.001
Race/Ethnicity, %			<0.001
Caucasian	69.2	73.7
African-American	9.0	5.4
Hispanic/Latino	15.7	12.7
Other	6.1	8.2
Female sex, %	33.4	36.8	0.01
MELD at offer, median (IQR)	22 (18, 25)	25 (22, 29)	<0.001
MELD at offer, %			<0.001
15–22	60.9	38.2
23–28	26.1	34.7
29–34	8.0	19.5
35–40	5.0	7.6
Indication for transplant, %			<0.001
NASH	10.7	15.4
HCV cirrhosis	31.2	14.9
HCC	14.3	15.9
Alcoholic liver disease	16.0	21.1
Other	27.7	32.7

Outcomes following OLD graft decline

To characterize outcomes among candidates who ever declined an OLD graft offer, we followed 23,709 candidates from the date of initial OLD graft decline until subsequent non-OLD graft LT, OLD graft LT, death, removal from the waitlist, or administrative censoring on 2/28/2018. We report five-year cumulative incidence of each potential outcome following the decision to decline the OLD graft accounting for competing events using the methods of Gooley et al (14). To determine whether outcomes following OLD graft decline varied by MELD at offer, we stratified this analysis by allocation MELD at offer, categorized as 15–24, 25–34, and 35–40. We compared MELD at offer and MELD at removal (last reported MELD) for each candidate using Wilcoxon matched-pairs signed-ranks test.

Creating matched populations

To quantify the potential survival benefit associated with OLD graft offer acceptance, we compared the overall mortality of 1,042 candidates who accepted and were transplanted with an OLD graft to 12,393 candidates who were matched on allocation MELD at offer (within 2 points) and declined the same OLD graft. Regression (as described below) was used to adjust for other clinical characteristics. A similar method has been used previously to quantify survival benefit of kidney offer acceptance (15). Candidates who accepted the OLD offer with a MELD below 35 were matched to those who declined the same offer within ±2 MELD points and also had MELD below 35. Candidates with MELD ≥35 were matched to those who declined the same offer within ±2 MELD points and also had MELD≥35 given significant differences in allocation priority for this subset of patients. We identified a median (interquartile range [IQR]) of 6 (2–16) candidates who had declined the same OLD graft with a similar MELD at offer for each candidate who accepted one. Candidates who accepted the initial OLD graft offer or had no available controls with MELD score within 2 points (n=269) were excluded from analysis as there were no candidates who declined the OLD to serve as matched controls.

Survival benefit of OLD graft acceptance

Candidates were followed from the date of OLD LT (if accepted OLD graft) or the date of OLD graft decline until the date of death or administrative censoring on 2/28/2018, irrespective of subsequent transplants. We used Cox regression with adjustment for candidate age at time of offer, race, biologic sex, and primary diagnosis for transplant (hepatitis C virus [HCV], hepatocellular carcinoma [HCC], non-alcoholic steatohepatitis [NASH], alcoholic liver disease, and other) to estimate survival benefit of accepting an OLD offer and adjust for clinical differences between patients who accepted versus declined the OLD graft. Candidates who declined an OLD graft offer but later accepted a different OLD graft were considered to have a time-varying exposure; in other words, these candidates were included in the OLD graft decline group from initial OLD graft decline until subsequent OLD graft LT (and censored then), and also included in the OLD graft acceptance risk group from the subsequent OLD graft LT until death or administrative censoring. Candidates who declined multiple OLD grafts were included as above (multiple truncated observations) for each OLD graft declined. We used clustering in the Cox regression model to identify cases and matched controls and to adjust standard errors.

To determine whether the survival benefit of accepting an OLD graft persisted across candidate phenotypes, we explored the association between accepting an OLD graft and mortality among candidates with age≥65, without exception points, who received local OLD liver offers, and finally, according to liver availability within the candidates’ donation service area (DSA). Liver availability was defined as the liver supply/candidate demand ratio within each DSA and categorized into quartiles. The supply/demand ratio was equal to the total number of livers recovered from within each DSA (that were used for transplant) divided by the total number of liver transplant candidates with MELD≥15 listed within each DSA. For each subgroup analysis, candidates were matched on the same OLD liver offer and MELD at offer (within 2 points).

Sensitivity analyses

To determine whether the use of candidates who declined multiple OLD offers as independently matched controls introduced bias in our final model, we performed a sensitivity analysis in which, for each candidate who accepted an OLD graft, we used random sampling to identify one candidate without replacement who declined the same graft and was matched within 2 MELD points.

To determine whether allowing candidates to have a time-varying exposure (i.e. decline and later accept an OLD graft) introduced bias, we performed a sensitivity analysis in which all candidates were followed from their initial decision to accept or decline the OLD offer until mortality or administrative censorship. Thus, we compared mortality of 722 candidates who accepted the first OLD graft they were offered to that of candidates who declined the same liver offers and were within 2 MELD points at the time of offer (the 589 candidates who declined and later accepted an OLD were followed from the date of initial decline only).

Finally, candidates who accepted the OLD offer from the top position of the match run (n=269) were excluded from the analyses as they had no available controls who declined the same OLD graft. To determine whether this exclusion introduced bias, we compared mortality between candidates who accepted the OLD graft offer from the top position on the match run with those who accepted it from a lower position and had available controls.

Statistical analyses

Confidence intervals were reported by the Louis and Zeger method (16). Analyses were two-tailed and α was set at 0.05. Analyses were performed using functions stcompet and clustering and stratification in the survival package stcox in Stata/SE, version 15 (StataCorp).

RESULTS

Study population

The median (IQR) age of OLDs was 74 (71–76) years and range of 70–90 years. Compared to candidates who declined OLD graft offers (n=23,709), candidates who accepted OLD grafts (n=1,311) were older (median 60.5 years vs. 57.8 years, p<0.001), more likely to be female (36.8% vs. 33.4%, p=0.01), and more likely to be Caucasian (73.7% vs. 69.2%, p<0.001) (Table 1). Candidates who accepted the OLD graft offers had a higher median MELD score at the time of offer (25 vs. 22, p<0.001), and were more likely to have NASH (15.4% vs. 10.7%), HCC (15.9% vs. 14.3%), or alcoholic liver disease (21.1% vs. 16.0%), but less likely to have HCV (14.9% vs. 31.2%).

Outcomes following OLD graft decline

Five years after their initial OLD graft decline, 51.2% of candidates had received a non-OLD graft LT, 2.5% had received an OLD graft LT, 24.5% had died without a transplant, 18.3% were removed from the waitlist, and 3.4% were still on the waitlist. However, these outcomes varied by MELD at the time of offer: 43.8% of MELD 15–22 candidates, 61.8% of MELD 23–28 candidates, 67.6% of MELD 29–34 candidates, and 63.9% of MELD 35–40 candidates received a non-OLD DDLT within five years of the initial OLD decline; 26.8% of MELD 15–22 candidates, 19.8% of MELD 23–28 candidates, 18.3% of MELD 29–34 candidates, and 29.3% of MELD 35–40 candidates died before receiving a deceased or living donor transplant; and 2.4% of MELD 15–22 candidates, 3.1% of MELD 22–28 candidates, 2.5% of MELD 29–34 candidates, and 1.0% of MELD 35–40 candidates eventually received an OLD DDLT five years after their initial OLD decline (Figure 1). Furthermore, 37.5% of all candidates who died without receiving a transplant between 12/28/2009–1/6/2017 were offered and declined at least one OLD graft before their date of death.

Figure 1: Outcomes among candidates who ever declined an older liver donor (OLD) graft between 12/28/2009–1/6/2017 (n= 23,709) stratified by allocation MELD at the time of first OLD decline. Candidates were followed forward from the date of initial OLD graft decline.

Five-year cumulative incidence is reported for each potential outcome in figure. OLD = older liver donor, DDLT= deceased donor liver transplant, *Reasons for removal from waitlist included declined transplant, transplanted at another center or in another country, candidate condition improved, living donor liver transplant, unable to contact candidate, and other

For the candidates who declined initial OLD graft offers and later underwent LT with a non-OLD graft, the median (IQR) donor age was 43 (28–54) years, 60.0% of the donors were male, 63.9% of the donors were Caucasian, and 18.8% of the donors were African American. Also, among the non-OLDs, 33.6% died of anoxia, 34.7% died of stroke, 6.3% were DCD, and the median (IQR) cold ischemia time was 6.0 (4.8–7.7) hours.

Change in MELD following OLD graft decline

Among the 60.9% of candidates (n=11,936) who declined an OLD graft and later received a non-OLD graft LT, median (IQR) MELD score increased from 22 (21–27) at initial OLD decline to 29 (25–33) at eventual non-OLD graft LT (p<0.001) (Table 2); median time to non-OLD graft LT among all candidates who declined an OLD graft was 2.76 years. Among the 2.5% of candidates (n=584) who declined an OLD graft and later underwent a subsequent OLD graft LT, median (IQR) MELD score increased from 22 (21–25) at initial OLD decline to 27 (22–29) at subsequent OLD graft LT (p<0.001) (Table 2). Among the 23.4% of candidates (n=5,543) who declined an OLD graft and died before receiving a transplant, median (IQR) MELD score increased from 22 (17–25) at OLD decline to 24 (17–33) at last recorded MELD (p<0.001) (Table 2).

Table 2. MELD score (median [IQR]) at date of first OLD decline and eventual outcome as of 2/28/2018 for 23,709 candidates.

Outcome	N	MELD at initial OLD decline	MELD at subsequent outcome	p value
Still Waiting	1,539	17 [16, 21]	15 [10, 19]	<0.001
Died	5,543	22 [17, 25]	24 [17, 33]	<0.001
Removeda	4,102	20 [17, 23]	18 [10, 25]	<0.001
OLD graft LT	589	22 [21, 25]	27 [22, 29]	<0.001
Non-OLD graft LT	11,936	22 [21, 27]	29 [25, 33]	<0.001

^aReasons for removal include transplanted in another country, unable to contact patient, candidate condition improved, and living donor liver transplant

Survival benefit of OLD graft acceptance

After matching on OLD graft and MELD at offer, candidates who accepted versus declined OLD graft offers had an overall five-year cumulative mortality of 23.4% versus 41.2% (p<0.001). After adjustment, candidates who accepted the OLD graft offer experienced nearly a 2-fold reduction in mortality compared to those who declined the same OLD graft offer (adjusted hazard ratio [aHR]: _0.450.52_0.59, p<0.001).

Candidates who accepted versus declined OLD graft offers had lower overall five-year cumulative incidence of mortality within each MELD stratum: 23.7% vs 45.2% for MELD 15–22, 26.6% vs 35.2% for MELD 23–28, 17.8% vs 29.3% for MELD 29–34, and 18.7% vs 42.2% for MELD 35–40 (all p<0.001) (Figure 2). After adjustment, accepting the OLD graft offer was associated with a 54%, 30%, 49%, and 76% reduction in mortality among candidates with MELD 15–22 (aHR: _0.380.46_0.55, p<0.001), MELD 23–28 (aHR: _0.570.70_0.86, p<0.001), MELD 29–34 (aHR: _0.350.51_0.74, p<0.001), and MELD 35–40 (aHR: _0.100.24_0.55, p=0.001) (Table 3).

Figure 2: Cumulative mortality after OLD offer acceptance and decline among candidates with (A) MELD 15–22, (B) MELD 23–28, (C) MELD 29–34, and (D) MELD 35–40 at time of offer.

Candidates matched on the liver offered and MELD score (within 2 points) at time of offer. Cumulative mortality post-decision to decline OLD offer includes candidates who subsequently underwent living donor LT, underwent younger donor deceased donor LT, were removed from the waitlist, and those who remained on the waitlist.

Table 3. Association between OLD-liver offer acceptance and mortality stratified by MELD at time of offer.

Survival benefit of OLD offer acceptance was observed in each MELD stratum, however, the survival benefit was attenuated slightly for candidates with MELD 23–28.

	aHR of OLD offer acceptance vs declinea	p value
MELD 15–22	0.380.460.55	<0.001
MELD 23–28	0.570.700.86b	<0.001
MELD 29–34	0.350.510.74	<0.001
MELD 35–40	0.100.240.55	0.001

^aCox regression model adjusted for candidate age, sex, race, diagnosis, and with an interaction between MELD at offer and OLD acceptance. Model accounts for clustering among accepting candidates and their matched controls. Candidates matched on the same liver offer and MELD (within 2 points) at time of offer.

^bAmong candidates with MELD 23–28, there was survival benefit associated with OLD offer acceptance (30% reduction in mortality), however, this survival benefit was statistically significantly less than that observed for candidates with MELD 15–22 and MELD 35–40 (p-interaction terms = 0.003 and 0.01, respectively).

Survival benefit of OLD graft acceptance within patient phenotypes

The survival benefit associated with OLD graft acceptance did not vary by candidate age (age<65 or ≥65) (p-interaction=0.9) (Table 4). After matching on exception point status, candidates with and without exception points at the time of offer experienced survival benefit with OLD acceptance, although the survival benefit was slightly attenuated among those with exception points (aHR among those without exception points: _0.300.37_0.46, aHR among those with exception points: _0.620.75_0.90, p-interaction<0.001). The survival benefit of OLD acceptance did not vary by local vs shared status of the OLD graft (p-interaction = 0.3), nor the liver availability of the DSA (p-interaction=0.4).

Table 4. Survival benefit of OLD-liver offer acceptance within candidate phenotypes.

Candidate subgroups	aHR of OLD offer acceptance vs declinea	p value	p value interaction
Candidate age			0.9
Age at offer < 65	0.460.530.63	<0.001
Age at offer ≥ 65	0.450.550.68	<0.001
Exception points at time of offerb			<0.001
No exception points	0.300.370.46	<0.001
Exception points	0.620.750.90	0.002
Shared status of liverc			0.3
Local OLD offer to candidates in same DSA	0.360.470.62	<0.001
Shared OLD offer to candidates in same DSA	0.460.550.67	<0.001
Liver availability in DSAd			0.4
Q1: Lowest supply/demand ratio DSAs	0.440.520.63	<0.001
Q2	0.400.510.65	<0.001
Q3	0.310.460.69	<0.001
Q4: Highest supply/demand ratio DSAs	0.330.751.71	0.5

OLD old liver donor; aHR adjusted hazard ratio; HCV Hepatitis C virus; DSA donation service area

^aCox regression models adjusted for candidate age, sex, race, diagnosis, and MELD at offer where appropriate. Models account for clustering among cases and matched controls. Candidates matched on the same liver offer and MELD (within 2 points) at time of offer.

^bCandidates matched on exception point status at the time of offer, allocation MELD at the time of offer, and OLD graft

^cCandidates matched on DSA, allocation MELD at offer, and OLD graft

^dLiver availability based on supply/demand ratio within each DSA. Supply/demand ratio calculated as crude ratio of livers recovered from within each DSA (that were used for transplant) divided by the total number of liver transplant candidates with MELD≥15 in each DSA; candidates matched on liver availability of their DSA, allocation MELD at offer, and OLD graft

Sensitivity analyses

Matching on OLD graft and MELD without replacement, inferences with respect to survival benefit of accepting an OLD graft were consistent with the main model (aHR of OLD acceptance: _0.490.59_0.71, p<0.001).

Following candidates from their first encounter with an OLD offer until death or administrative censoring, without time-varying exposure of OLD acceptance, inferences were consistent with the main model (aHR of OLD acceptance _0.410.50_0.59, p<0.001).

Among the 269 candidates for whom we could not directly identify matched controls because they accepted the OLD offer from the top position of the match run, survival benefit was seen when compared to candidates who had declined different OLD grafts (30% reduction in mortality; aHR OLD acceptance: _0.550.70_0.88, p=0.002), but was less than that observed among candidates who had accepted an OLD graft from farther down on the match run (aHR: _0.460.52_0.60, p<0.001). The median MELD at OLD offer acceptance was 28 (IQR 23–34) among those who accepted from the top position versus 25 (IQR 21–28) among those who accepted an OLD graft from farther down on the match run. In a model that included all candidates who had accepted an OLD graft (from any position), inferences were consistent with the main model (aHR of OLD acceptance: _0.490.56_0.62, p<0.001).

DISCUSSION

In this national study of 1,311 OLD graft recipients and 23,709 candidates who declined the same OLD grafts, we found accepting OLD grafts was associated with a substantial survival benefit. This was seen across all MELD categories, with greater mortality reduction at the highest MELDs: those with MELD 15–22 had a 54% reduced mortality, those with MELD 23–28 had a 30% reduced mortality, those with MELD 29–34 had a 49% reduced mortality, and those with MELD 35–40 had a 76% reduced mortality by accepting an OLD graft. Among candidates who declined OLD graft offers, 24.5% died before ever receiving a transplant. After OLD graft offer decline, candidates’ median MELD scores significantly increased between the time of next offer and subsequent transplant (OLD graft LT or non-OLD graft LT). Importantly, we found that only candidates without HCV derived a survival benefit from OLD graft acceptance; those with HCV as a primary indication for transplantation, on average, did not derive a survival benefit.

Our finding that a high percentage of candidates died (37.5%) after receiving at least one OLD graft liver offer is consistent with findings that demonstrated 84% of patients who died or were removed from the waitlist with MELD>15 had declined at least one liver offer (3). In our study, it is important to note the proportion that were eventually transplanted in each MELD category as well; only 43.8%, 61.8%, 67.6%, and 63.9% of those who declined an OLD graft offer in each MELD category (MELD 15–22, 23–28, 29–34, 35–40) subsequently underwent a non-OLD graft LT within 5 years. As a result, 26.8%, 19.8%, 18.3%, and 29.3% of those who declined an OLD graft offer eventually died without a transplant in each MELD category (MELD 15–22, 23–28, 29–34, 35–40). Finally, our finding of five-year mortality for OLD graft LT recipients of 23.4% is consistent with national (5, 12, 17, 18) and single center (19) reports of post-LT survival for OLD recipients.

Our finding of survival benefit in high MELD LT candidates associated with OLD offer acceptance might seem discordant with reports that report high MELD candidates have poor post-transplant outcomes with OLD grafts (20–23). However, these studies did not quantify survival benefit of LT in high MELD candidates compared to the counterfactual scenario of declining the OLD graft. The possibility of better post-transplant survival with a potential future offer from a younger donor must be weighed against the chance of not receiving another offer after OLD graft decline and dying on the waitlist. Our findings must be taken in the context of the evaluation of centers’ post-transplant outcomes. Thus, decline of an OLD graft offer will not affect centers’ post-transplant outcomes, but acceptance of an OLD graft offer may worsen outcomes even though these organs provide a survival benefit.

Our finding that acceptance of an OLD graft offer in candidates, across all ages, was associated with a survival benefit over declining an OLD graft offer is discordant a report that older age (age≥60) is associated with poor outcomes after LT with an OLD graft (5). Again however, this study did not quantify survival benefit of OLD graft acceptance which is a different question than post-LT outcomes. Thus, older candidates derive a survival benefit from OLD graft offer acceptance and should not be ruled out as potential recipients of these grafts, especially as outcome in older LT recipients improve (24).

One notable limitation to our study is that we did not have information on who made the decision to decline an OLD graft offer. Also, it is possible that center-level differences in OLD graft utilization led to sweeping declines across entire centers (rather than individual patient-level decisions). It is important to note that each liver was “transplantable,” i.e. eventually used by some center/recipient for transplantation. Also, it is important to note that increased acceptance of the organs studied in this analysis would not increase the volume of transplants performed, as we only studied livers that were eventually accepted by someone, and extending our findings to OLD grafts that were declined will require clinical judgment. Given the high discard of OLD grafts (5), it is likely that some of these organs might have provided survival benefit to select candidates; further work is needed to explore the potential for increased utilization of OLD grafts.

In conclusion, acceptance of an OLD graft confers survival benefit for waitlist candidates across all MELD scores, especially in high MELD candidates. Notably, both older and younger candidates derived a survival benefit from OLD graft acceptance. Patients and providers should carefully weigh the consequences of declining an OLD graft offer, as one-fourth of candidates die after such a decline. These findings can guide future clinical decision making for transplant providers and improve patient counseling in considering offers.

Abbreviations:

aHR

adjusted hazard ratio

BMI

body mass index

CIT

cold ischemia time

ESLD

end-stage liver disease

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

HRSA

Health Resources and Services Administration

IQR

interquartile range

LT

liver transplantation

MELD

Model for End-stage Liver Disease

NASH

non-alcoholic steatohepatitis

OLD

older liver donor

OPTN

Organ Procurement and Transplantation Network

SRTR

Scientific Registry of Transplant Recipients