Impact of Older Donor Age on Recipient and Graft Survival After LDLT: The US Experience : Transplantation

Journal Logo

Original Clinical Science—Liver

Impact of Older Donor Age on Recipient and Graft Survival After LDLT: The US Experience

Kanneganti, Mounika MD1; Olthoff, Kim M. MD2; Bittermann, Therese MD, MSCE3,4

Author Information
Transplantation 107(1):p 162-171, January 2023. | DOI: 10.1097/TP.0000000000004289
  • Free
  • SDC
  • Infographic

Abstract

INTRODUCTION

As donor age increases, the liver allograft’s ability to overcome certain stressors (eg, ischemia) may lessen, risking unsatisfactory outcomes.1 Although extensively studied in the context of deceased donor liver transplantation (DDLT), data are conflicting regarding the independent contribution of increasing donor age on recipient and graft survival, as there are undoubtedly many other factors involved.1-4 Nevertheless, given the ongoing donor organ shortage, US transplant centers are increasingly accepting older donor liver allografts for their recipients, even if otherwise highly selected.5,6 In 2021, more than one-third of all liver transplants (LTs) in the United States were performed using allografts procured from donors ≥50 y.7

As centers work to increase living donor liver transplantation (LDLT) in the United States, the issue of increasing donor age brings up unique considerations. These include donor safety and the potential for impaired liver regeneration in both the donor and recipient. In contrast, the relevance of increasing donor age on the allograft’s resilience to ischemia-reperfusion injury is a lesser concern in LDLT.4 Studies from Asia, where LDLT is highly prevalent, have shown inconsistent effects of increasing living donor (LD) age on recipient and outcomes.8-12 Mixed findings have also been described with regard to donor outcomes.13,14 For example, a prior study from Turkey demonstrated unsatisfactory donor complication rates among those ≥50 y with remnant liver volume <35%.15 In contrast, another study from Canada showed no difference in outcomes between older and younger donors, though the age cut-point was lower (44 y).16

Data on the association of increasing donor age on LDLT outcomes in the United States have been primarily obtained from the Adult-to-Adult Living Donor Liver Transplantation Studies (A2ALL).17,18 A2ALL studies demonstrated that increasing donor age was associated with early allograft dysfunction and higher recipient mortality.19,20 However, no relationship between donor age and donor postoperative complications was identified.17 Notably, the A2ALL studies were conducted during a time when LDLT center selection practices were conservative, and thus the use of LDs >50 y was rare.18 With centers now expanding their donor acceptance criteria, we hypothesized that US trends have changed since A2ALL, potentially leading to increased consideration of older donors.

Therefore, the objectives of this study were to (1) assess trends in the use of allografts procured from older living liver donors in the United States, (2) evaluate the relationship between older LD age and recipient outcomes, (3) compare the impact of increasing donor age on LDLT versus DDLT recipients, and (4) ascertain the safety of LD surgery with increasing donor age.

MATERIALS AND METHODS

This was a retrospective cohort study of LDLTs from January 01, 2005 to December 31, 2019 using national data from the United Network for Organ Sharing (UNOS). Multiorgan transplants and recipients of prior LT were excluded. A comparator group of adult DDLT recipients during the same period was also identified, which additionally excluded Status 1 (ie, emergent LT) recipients and transplants with donor allografts aged ≥65 y (to allow for comparable donor age between the LDLT and DDLT groups).

Exposures and Other Study Variables

Our primary exposure was LD age, which was evaluated in multiple forms. In descriptive analyses, the following categories were used: <40 y, 40–49 y, 50–55 y, and >55 y. In outcome analyses, we evaluated donor age continuously (per decile) and in categorical form by 10-y increments (18–29 y, 30–39 y, 40–49 y, and ≥50 y). We distinguished the 50–55 y versus >55 y group in the descriptive analyses as 55 y is an upper age limit for living donation at multiple US centers. In multivariable analyses, however, the >55 y stratum was small, producing point estimates with very low precision, and thus using a decile approach was more practical.

Other LD variables included age, sex, race/ethnicity, relationship to liver recipient, history of donor liver biopsy, body mass index (BMI), hepatic steatosis, lobe donated, cancer history, and tobacco use. Predonation liver volume was estimated using: (0.72√[body surface area] + 0.171).3,18 Preoperative laboratory parameters included alanine aminotransferase, total bilirubin, alkaline phosphatase, albumin, international normalized ratio, creatinine, and estimated glomerular filtration rate estimated using the 4-variable Modification of Diet in Renal Disease equation.21 Recipient characteristics at LT included sex, age, race/ethnicity, etiology of liver disease, waiting time, Model for End-stage Liver Disease (MELD) score, ascites severity (none, mild, and moderate–severe), and hepatic encephalopathy severity (none, grades 1–2, grades 3–4).

Outcomes

The primary study outcomes were recipient and graft survival. Donor outcomes included: (i) early (≤6 wk) postoperative complications and (ii) laboratory values and complications at 6 mo, 1 y‚ and 2 y.

Statistical Analyses

Baseline recipient and donor characteristics were compared according to donor age groups using chi-squared tests for categorical variables and Kruskal Wallis tests for continuous variables. Temporal trends in LDLTs performed by age category were assessed. Center variability in LDLT practice by donor age category was investigated among centers with ≥15 LDLTs over the study period (ie, ≥1 LDLT per y).

Time-to-event analyses evaluated the relationship between LD age and recipient patient and graft survival. Separate analyses evaluated donor age in continuous and categorical forms (both per decile), as it was hypothesized that the relationship between donor age and outcomes was not linear. Kaplan-Meier curves for each outcome were drawn with survival functions compared using the log-rank test.

Multivariable Cox proportional hazards models evaluated the association of donor age on recipient and graft survival, adjusted for: recipient age, recipient sex, recipient race/ethnicity, native MELD score at LT, liver disease etiology, ascites and encephalopathy severity, donor liver lobe type, donor–recipient sex mismatch, donor–recipient race/ethnicity mismatch, calendar year‚ and center LDLT volume during the study period. For each recipient outcome, the following interactions with donor age were assessed: recipient age, native MELD score, ascites severity, and donor liver lobe received. To determine whether the impact of increasing donor age on outcomes changed over time, the interactions of donor age with LT year and with LT era (2005–2009, 2010–2014, and 2015–2019) were evaluated.

The effect of donor age on recipient and graft survival was then compared between DDLT and LDLT recipients using stratified multivariable Cox models. The significance of the interaction between donor type and continuous donor age was evaluated. As overall LDLT and DDLT outcomes have improved over time and to ensure that the results were most relevant to current day practice, the study period was restricted to the most recent 10 y (ie, 2010–2019). These models were adjusted for recipient age, recipient sex, recipient race/ethnicity, native MELD score at LT, liver disease etiology, ascites, and encephalopathy severity. Donor age was again evaluated as a continuous (per 10-y increase) and as a categorical variable (18–29, 30–39, 40–49, and ≥50 y).

Lastly, donor complications in the first 6 wk after LDLT were reported according to donor age category (<50, 50–55, and >55 y) using descriptive statistics. Donor-related adverse events and postoperative laboratory parameters at 6 mo, 1 y, and 2 y from liver donation were compared by donor age. Data missingness was also evaluated and reported.

This study was approved by the Institutional Review Board of the University of Pennsylvania. All analyses were performed using STATA version 13 (College Station, TX).

RESULTS

Recipient and Donor Characteristics by LDLT Donor Age Category

The final study cohort included 3539 LDLT recipients and their 3539 donors at 65 transplant centers in the United States. The number of LDLTs performed by donor age category were as follows: age <40 (N = 2111), age 40–49 (N = 910), age 50–55 (N = 420), and age >55 (N = 98). The oldest living liver donor in the cohort was 63 y. The overall recipient cohort was 54.9% male, 82.5% White, with a median age of 55 y (interquartile range [IQR] 46–62) and median native MELD score of 15 (IQR 11–19; Table 1). Liver disease etiology was: 23.1% hepatitis C virus, 20.5% nonalcoholic steatohepatitis, 17.8% primary sclerosing cholangitis, and 12.3% alcohol-related liver disease. Additionally, 18.4% of the overall cohort had moderate to severe ascites, and 4.1% had grade 3–4 hepatic encephalopathy. There were few meaningful differences in recipient characteristics according to donor age (Table 1). For example, there was no difference in median recipient age at LT (P = 0.159) or in the degree of portal hypertensive complications (P = 0.450 for ascites and P = 0.340 for encephalopathy). There was also no difference in waiting time for patients transplanted with an older graft (P = 0.838).

TABLE 1. - Recipient characteristics by donor age categorya
Age <40 (N = 2111) Age 40–49 (N = 910) Age 50–55 (N = 420) Age >55 (N = 98) P
Male, n (%) 1166 (55.2) 478 (52.5) 244 (58.1) 54 (55.1) 0.273
Recipient age, median (IQR), y 56 (46–62) 53 (45–64) 54 (48–59) 57 (51–61) 0.159
Race/ethnicity,n(%) <0.001
 White 1682 (79.7) 780 (85.7) 375 (89.3) 84 (85.7)
 Black 70 (3.3) 29 (3.2) 13 (3.1) 1 (1.0)
 Hispanic 268 (12.7) 79 (8.7) 20 (4.8) 10 (10.2)
 Asian 65 (3.1) 20 (2.2) 7 (1.7) 3 (3.1)
 Other 26 (1.2) 2 (0.2) 5 (1.2) 0 (0.0)
Diagnosis, n (%) 0.189
 HCV 499 (23.6) 186 (20.4) 105 (25.0) 27 (27.6)
 Alcohol 244 (11.6) 112 (12.3) 62 (14.8) 18 (18.4)
 HBV 43 (2.0) 12 (1.3) 7 (1.7) 1 (1.0)
 NASH/cryptogenic 435 (20.6) 204 (22.4) 69 (16.4) 17 (17.3)
 PSC 367 (17.4) 164 (18.0) 78 (18.6) 20 (20.4)
 Autoimmune 109 (5.2) 50 (5.5) 18 (4.3) 5 (5.1)
 Other 237 (11.2) 110 (12.1) 51 (12.1) 8 (8.2)
 PBC 177 (8.4) 72 (7.9) 30 (7.1) 2 (2.0)
Calculated BMI at LT,median (IQR), kg/m2 26 (23,30) 26 (23,30) 26 (23,29) 27 (24, 29) 0.158
Lab MELD at LT, median (IQR) 15 (11,19) 15 (11,19) 15 (11,19) 16 (11, 21) 0.453
Ascites at LT, n (%) 0.450
 Absent 710 (33.8) 322 (35.7) 146 (35.4) 31 (32.0)
 Mild 992 (47.2) 426 (47.2) 196 (47.5) 41 (42.3)
 Moderate–severe 400 (19.0) 155 (17.7) 71 (17.2) 25 (25.8)
Encephalopathy at LT, n (%) 0.340
 None 1006 (47.9) 443 (49.0) 185 (44.8) 44 (45.4)
 Grade 1–2 1009 (48.0) 418 (46.2) 217 (52.5) 48 (49.5)
 Grade 3–4 87 (4.1) 43 (4.8) 11 (2.7) 5 (5.2)
Waitlist time, median (IQR), d 147 (74–307) 151 (75–307) 146 (82–321) 171 (80–388) 0.838
aPercent missingness for recipient characteristics across all ages was <1%.
BMI, body mass index; HBV, hepatitis B virus; IQR, interquartile range; LT, liver transplantation; MELD, Model for End-Stage Liver Disease; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis.

The overall donor cohort was 47.8% male and 82.8% White (Table 2). Right lobe donors were less frequent among donors >55 y, though this did not reach statistical significance (P = 0.120). Donor age was associated with a preoperative liver biopsy practice (P = 0.013), although there was no clear trend. Among donors undergoing a preoperative liver biopsy (n = 1001), ≥5% steatosis (macro- or micro-) was found in 20.6% of donors aged <40 y, 32.2% aged 40–49 y, 30.6% aged 50–55 y, and 22.7% aged >55 y (P = 0.002). The prevalence of obese donors (ie, with BMI ≥30 kg/m2) was different by donor age category, with donors aged 40–49 having the highest prevalence (18.5%, P < 0.001). A marginally higher rate of prior malignancy was observed among donors aged 50–55 and >55 (P = 0.001). There was no clear trend between increasing donor age and estimated predonation liver volume (P = 0.093). Donor relationship differed by donor age (P < 0.001): 62.2% of donors >55 y were nonbiological‚ whereas 60.6% of donors <40 y were first-degree relatives.

TABLE 2. - Donor characteristics by donor age categorya
Age <40 (N = 2111) Age 40–49 (N = 910) Age 50–55 (N = 420) Age >55 (N = 98) P
Male, n (%) 1104 (52.3) 378 (41.5) 166 (39.5) 44 (44.9) <0.001
Race, n (%) <0.001
 White 1681 (79.6) 780 (85.7) 377 (89.8) 91 (92.9)
 Black 70 (3.3) 28 (3.1) 9 (2.1) 1 (1.0)
 Hispanic 252 (11.9) 80 (8.8) 23 (5.5) 5 (5.1)
 Asian 59 (2.8) 19 (2.1) 8 (1.9) 1 (1.0)
 Other 49 (2.3) 3 (0.3) 3 (0.7) 0 (0)
Donor liver lobe, n (%) 0.120
 Left lateral segment 16 (0.8) 8 (0.9) 0 (0.0) 2 (2.0)
 Left lobe 296 (14.0) 113 (12.4) 48 (11.4) 17 (17.3)
 Right lobe 1799 (85.2) 789 (86.7) 372 (88.6) 79 (80.6)
Liver donation relation, n (%) <0.001
 First-degree relative 1279 (60.6) 427 (46.9) 169 (40.2) 33 (33.7)
 Other biological relative 231 (10.9) 98 (10.8) 29 (6.9) 4 (4.1)
 Nonbiological related 227 (10.8) 160 (17.6) 94 (22.4) 35 (35.7)
 Nonbiological unrelated 374 (17.7) 225 (24.7) 128 (30.5) 26 (26.5)
Donor BMI category, n (%) <0.001
 <18.5 13 (0.6) 3 (0.4) 2 (0.5) 0 (0.0)
 18.5–24.9 831 (41.0) 270 (31.3) 129 (32.2) 38 (39.6)
 25.0–29.9 872 (43.0) 431 (49.9) 212 (53.0) 43 (44.8)
 [scolor_start FADADD]≥[/scolor]30.0 313 (15.4) 160 (18.5) 57 (14.2) 15 (15.6)
Predonation liver volume, median (IQR), L b 1.56 (1.4–1.7) 1.55 (1.4–1.7) 1.51 (1.4–1.7) 1.56 (1.4–1.7) 0.093
Biopsy of donor, n (%) 562 (26.6) 283 (31.1) 134 (31.9) 22 (22.5) 0.013
Macrovesicular steatosis, n (%) c 0.419
 <5% 289 (51.4) 137 (48.4) 68 (50.7) 7 (31.8)
 5–9% 60 (10.7) 37 (13.1) 16 (11.9) 3 (13.6)
 10–19% 29 (5.2) 20 (7.1) 13 (9.7) 1 (4.5)
[scolor_start FADADD]≥[/scolor]20% 5 (0.9) 5 (1.8) 2 (1.5) 0 (0.0)
 Missing 179 (31.9) 84 (29.7) 35 (26.1) 11 (50.0)
Microvesicular steatosis, n (%) c 0.070
 <5% 306 (54.5) 142 (50.2) 70 (52.2) 10 (45.5)
 5–9% 12 (2.1) 13 (4.6) 8 (6.0) 1 (4.5)
 10–19% 7 (1.3) 11 (3.9) 7 (5.2) 0 (0.0)
[scolor_start FADADD]≥[/scolor]20% 2 (0.4) 2 (0.7) 1 (0.7) 0 (0.0)
 Missing 235 (41.8) 115 (40.6) 48 (35.8) 11 (50.0)
Donor cancer history, n (%) 0.001
 No cancer history 2085 (98.8) 889 (97.7) 402 (95.7) 94 (95.9)
 Prior skin cancer 5 (0.2) 11 (1.2) 9 (2.1) 2 (2.0)
 Prior non–skin cancer 21 (1.0) 10 (1.1) 9 (2.1) 2 (2.0)
ALT, median (IQR) 20 (15–29) 21 (15–28) 21 (16–28) 21 (18–28) 0.411
Total bilirubin, median (IQR) 0.6 (0.4–0.8) 0.6 (0.4–0.8) 0.5 (0.4–0.7) 0.6 (0.4–0.8) 0.192
Alkaline phosphatase, median (IQR) 63 (53–75) 63 (51–76) 67 (57–80) 66 (58–81) <0.001
Serum albumin, median (IQR) 4.4 (4.2–4.7) 4.3 (4.1–4.6) 4.4 (4.2–4.6) 4 (3.8–4.3) <0.001
INR, median (IQR) 1 (1.0–1.1) 1 (1.0–1.1) 1 (1.0–1.0) 1(0.9–1.0) <0.001
Serum Cr, median (IQR) 0.8 (0.7–1.0) 0.8 (0.7–1.0) 0.85 (0.7–1) 0.87(0.73–1) 0.845
eGFR, median (IQR) d 99.9 (88.2–113.3) 87.7 (78.7–98.3) 83 (74–94) 82 (76–92) <0.001
Tobacco use, n (%) 597 (29.7) 261 (30.3) 120 (30.3) 29 (30.9) 0.984
aUnless otherwise noted in the table, percent missingness for donor characteristics across all ages: tobacco 5.0%, BMI 4%, laboratory values 1–3%, all other variables <1%.
bCalculated using (0.72√[body surface area] + 0.171).3
cAmong donors with a preoperative liver biopsy (n = 1001).
dCalculated using Modified Diet in Renal Disease 4-variable equation.
ALT, alanine aminotransferase; BMI, body mass index; Cr, creatinine; eGFR, estimated glomerular filtration rate; INR, international normalized ratio; IQR, interquartile range.

Temporal Trends in LDLT by Donor Age Category

Although there was an increase in the absolute number of LDLTs across all donor age groups from 2011 onward, the proportion of donors aged 40–49 y, 50–55 y, and >55 y remained relatively stable (Figure 1A). Overall median LD age was 38 y (IQR: 30–46) in 2005 and 37 y (IQR: 30–46) in 2019, with little variability in between. In 2019, 26.0% of LDs were 40–49 y, 13.1% were aged 50–55 y, and 2.5% were >55 y in the United States.

F1
FIGURE 1.:
Temporal (A) and center (B) trends in LDLT by donor age. For (B), each vertical bar/column represents 1 center with centers ranked by increasing LDLT volume. Center trends graph includes N = 33 centers performing ≥15 LDLTs during the study period (2005–2019) or an average of ≥1 LDLT per y. LDLT, living donor liver transplantation.

Center Variability in LDLT With Older Donors

The overall median center volume from 2005 to 2019 was 15 LDLTs (IQR: 4–81; n = 65 centers). There was no difference in center LDLT volume by donor age category (P = 0.786). Center trends in LD age are shown among the 33 centers performing ≥15 LDLTs over the 15-y study period (Figure 1B). Although only 23 (69.7%) centers performed ≥1 LDLT with an LD aged >55 y, all performed ≥1 with a donor aged 40–55 y. Notably, although there was a trend toward centers with higher LDLT volume performing more LDLTs aged >55 y, the largest center (n = 396 LDLTs) had no donors >55 y. Median LD age by center ranged from 29 to 44 y (Figure S1, SDC, https://links.lww.com/TP/C514).

Age and LDLT Recipient and Graft Survival

One-year LDLT recipient mortality was 5.8%, 8.0%, 9.4%, and 11.1% for donor age 18–29, 30–39, 40–49, and ≥50 y, respectively (P = 0.007), whereas the 1-y retransplant rate was 4.5%, 6.7%, 7.1%, and 8.8%, respectively, by age group (P = 0.021). Median recipient length of stay post-LDLT was not different according to donor age (P = 0.916).

The unadjusted hazard ratio (HR) for recipient survival was 1.16 per 10-y increase in donor age (95% confidence interval [CI], 1.08-1.25; P < 0.001), whereas the unadjusted HR for graft survival was 1.18 (95% CI, 1.11-1.26; P < 0.001; Table S1, SDC, https://links.lww.com/TP/C515). Unadjusted Kaplan-Meier curves by donor age are shown in Figure 2Aand C, respectively. Five-year recipient and graft survival with LD age ≥50 y were 79.7% and 71.4%, whereas at 10 y posttransplant rates were 67.4% and 58.6%, respectively.

F2
FIGURE 2.:
Unadjusted and adjusted patient survival (A and B, respectively) and graft survival (C and D, respectively) post-LDLT by donor age. LDLT, living donor liver transplantation; LT, liver transplantation.

In multivariable analyses, donor age remained independently associated with recipient and graft survival (Table 3). Adjusted HRs for recipient and graft survival were 1.16 (95% CI, 1.07-1.25; P < 0.001) and 1.18 (95% CI, 1.05-1.26; P < 0.001) per 10-y increase in donor age, respectively. The adjusted HRs for patient survival and graft survival for donors aged ≥50 y were 1.49 (P = 0.012) and 1.61 (P < 0.001), respectively, relative to a donor aged 18–29 y (adjusted recipient/graft survival curves in Figure 2B and D, respectively). There was no significant interaction between donor age and recipient age, MELD, ascites‚ or lobe type for patient and graft survival models (for recipient survival P = 0.386, 0.921, 0.810, 0.769; for graft survival P = 0.987, 0.774, 0.995, 0.695, respectively). There was also no significant interaction between donor age and LT year (P = 0.289 for recipient survival and P = 0.482 for graft survival) or LT era (P = 0.753 and P = 0.689, respectively), indicating that there has been no change in the association of donor age on recipient LDLT outcomes over time.

TABLE 3. - Association of living donor age with adjusted recipient and graft survival for living donor liver transplantations performed between 2005 and 2019 (N = 3507)
Recipient survival Graft survival
Adjusted HR (95% CI) P Adjusted HR (95% CI) P
Living donor age, per 10-y increase 1.16 (1.07-1.25) <0.001 1.18 (1.05-1.26) <0.001
Living donor age 0.012 <0.001
 18–29 Reference Reference
 30–39 1.18 (0.95-1.46) 1.29 (1.07-1.55)
 40–49 1.31 (1.05-1.64) 1.38 (1.14-1.67)
 ≥50 1.49 (1.16-1.92) 1.61 (1.30-1.99)
Multivariable models adjusted for recipient age, recipient sex, recipient race/ethnicity, native Model for End-Stage Liver Disease score at liver transplantation, liver disease etiology, ascites and encephalopathy severity, donor liver lobe type, donor–recipient sex mismatch, donor–recipient race/ethnicity mismatch, calendar y, and center living donor liver transplantation volume during the study period.
CI, confidence interval; HR, hazard ratio.

The Impact of Older Donor Age on Outcomes After LDLT Versus DDLT

The impact of increasing donor age was compared between 2570 LDLTs and 46 259 DDLTs performed from 2010 to 2019. In stratified adjusted analyses, increasing donor age had a lesser impact on recipient survival after DDLT than after LDLT: HR 1.06 (95% CI, 1.04-1.08; P < 0.001) per 10-y increase for DDLT and 1.16 (95% CI, 1.04-1.29; P = 0.007), respectively (Table 4; P = 0.058 for the interaction of donor type and donor age). Similar results were also observed for graft survival: 1.08 (95% CI, 1.06-1.09; P < 0.001) per 10-y increase after DDLT and 1.20 (95% CI: 1.10-1.31; P < 0.001; P = 0.019 for interaction) after LDLT. Compared with donor age 18–29 y, donor age ≥50 y was associated with an adjusted HR for patient survival of 1.18 (95% CI, 1.11-1.25; P < 0.001) for DDLT recipients and 1.52 (95% CI, 1.07-2.17; P = 0.020) for LDLT recipients. The adjusted HR for graft survival for donors ≥50 y was 1.25 (95% CI, 1.18-1.32; P < 0.001) after DDLT and 1.74 (95% CI, 1.10-1.80; P < 0.001) after LDLT, respectively (Table 4).

TABLE 4. - Association of donor age with adjusted recipient and graft survival among LDLT and DDLT recipients from 2010 and 2019
LDLT (N = 2563) DDLT (N = 45 954)
Adjusted HR (95% CI) P Adjusted HR (95% CI) P
Recipient survival
 Living donor age, per 10-y increase 1.16 (1.04-1.29) 0.007 1.06 (1.04-1.08) <0.001
 Living donor age 0.103 <0.001
  18–29 Reference Reference
  30–39 1.33 (0.99-1.79) 1.01 (0.95-1.08)
  40–49 1.31 (0.96-1.80) 1.04 (0.98-1.11)
  ≥50 1.52 (1.07-2.17) 1.18 (1.11-1.25)
Graft survival
 Living donor age, per 10-y increase 1.20 (1.10-1.31) <0.001 1.08 (1.06-1.09) <0.001
 Living donor age 0.002 <0.001
  18–29 Reference Reference
  30–39 1.41 (1.10-1.80) 1.03 (0.97-1.10)
  40–49 1.43 (1.10-1.85) 1.09 (1.03-1.16)
  ≥50 1.74 (1.30-2.33) 1.25 (1.18-1.32)
The model is adjusted for recipient age, recipient sex, recipient race/ethnicity, native Model for End-Stage Liver Disease score at liver transplantation, liver disease etiology, ascites, and encephalopathy severity. The P value for interaction of donor type and continuous donor age was 0.058 for recipient survival model and 0.019 for graft survival model.
CI, confidence interval; DDLT, deceased donor liver transplantation; HR, hazard ratio; LDLT, living donor liver transplantation.

Postoperative Donor Complications

There was no clear association between increasing donor age and early (≤6 wk) donor postoperative complications (Figure 3). Early (≤6 wk) biliary complications were numerically higher among donors >55, but not statistically significant (7.1% versus 3.1% for donors <40 y; P = 0.156). Rates of adverse events and postoperative complications at 6-mo, 1-y, and 2-y are shown in Figure 4Aand B and Figure S2, SDC, https://links.lww.com/TP/C515. Again, no clear differences were observed by donor age, although missingness was significant particularly at 1-y and 2-y.

F3
FIGURE 3.:
Early (≤6 wk) postoperative donor complications by age group. Missingness <1% for all outcomes.
F4
FIGURE 4.:
Postoperative donor adverse events (A) and specific complications (B) at 6 mo by age group. (A) Messiness for ER visits: 57.5%; for readmissions: 8.1%; for all complications: 8.1 %. (B) Messiness for each complication: 8.1%. ER, emergency room.

There were no meaningful differences in postoperative donor laboratory parameters by donor age (Table S2, SDC, https://links.lww.com/TP/C515). As shown in Figure S3, SDC, https://links.lww.com/TP/C515, there was considerable missingness observed for postoperative donor laboratory parameters with up to 60%–70% missing at 2 y post-LDLT. However, missingness was consistently lower among donors aged >55 y (Figure S4, SDC, https://links.lww.com/TP/C515).

DISCUSSION

In the last decade, there has been an increased interest in LDLT in the United States. However, overall center acceptance of older living liver donors has not markedly changed. And, though there were differences among centers, these were relatively small and all centers pursued few LDLTs with LDs ≥50 y. Although select studies from Asia have demonstrated that it is safe to pursue LDLT in carefully selected older living liver donors, even when aged >60 y, there is still significant hesitancy at a number of experienced Asian LDLT centers.22-24 Among US LDLT recipients, we find that the negative impact of increasing donor age had important effects on graft outcomes, with nearly 1 in 10 recipients of LDs aged ≥50 y undergoing retransplantation by the first year. Nevertheless, long-term outcomes remained largely acceptable with LD age ≥50 y with 5- and 10-y graft survival reaching 71.4% and 58.6%, respectively. Interestingly, the negative effect of increasing donor age on graft survival was found to be more profound in the setting of LDLT compared with DDLT. Thus, although LDLT is frequently believed to afford better long-term outcomes over DDLT, our results suggest that this advantage may be lost in the setting of older LD age. The decision to pursue LDLT with an older donor must therefore take into consideration the recipients’ unique clinical circumstances, their potential availability of alternate donors‚ and, in particular, their risk of death on the transplant waitlist. Further studies are needed to assess whether a survival benefit of early LDLT with an older donor is nevertheless advantageous over continuing to wait for an alternate LD or DDLT.

We noted an absence of consistent recipient and donor selection practices in the setting of older LDs. For example, centers did not preferentially perform LDLT with older donors in older recipients or in those with greater severity of illness. There was also no difference in the prevalence of overweight or obese donors with increasing age. Although older LD age did not influence the decision to pursue a preoperative liver biopsy, we did observe a trend toward less steatosis among LDs >55 y. As restricting the cohort to donors with a liver biopsy would inherently cause selection bias, this could not be evaluated further in our analyses. However, it is possible that the negative impact of increasing age could be more profound considering centers may have been more selective in using these allografts. Future studies leveraging noninvasive steatosis measurements (eg, MR spectroscopy) could also evaluate steatosis as an effect modifier of increasing donor age on outcomes.25-27 Interestingly, there were differences in the donor–recipient relationship noted. Donors >55 y were more often nonbiological, potentially reflecting the unavailability or unsuitability of younger biologically-related donors.

In this study, we found that early donor postoperative complications and adverse events were not impacted by donor age. Early (≤6 wk) donor postoperative data in UNOS were largely complete (<1% missing). Early biliary complications were more common among donors aged >55 y, though this did not reach statistical significance, likely due to inadequate power. The evaluation of donor outcomes beyond the early postoperative period was unfortunately significantly limited by data missingness, which became increasingly problematic with longer follow-up duration reaching 50% for event rates and 60%–70% for laboratory parameters at 2-y in some instances. There was no indication that data missingness improved over time. We did, however, find lower rates of missing data with increasing age, as it is likely that older donors are more accustomed to interacting with the health system on a regular basis compared with young, otherwise healthy donors. Nevertheless, the issue of maintaining LD engagement and follow-up with the LDLT team beyond the early postoperative recovery period remains an ongoing challenge at many centers. Additional studies are needed to more comprehensively determine whether liver donation at an older age is associated with potential long-term health effects.

Since the end of the A2ALL study cohort, the US LDLT community has been in great need of updated donor outcome data on a large scale. This is critical for donor education during the evaluation process and also for LDLT team decision-making when approving potential donor candidates. Many adverse event rates that are discussed with potential donors are still directly quoted from the seminal A2ALL studies, yet there are concerns that these do not accurately reflect the current era; an era with overall more LDLT experience, but also with more centers performing LDLT and also experienced centers expanding their donor selection criteria. One inherent challenge with UNOS data is that centers may be less inclined to directly report complications. In the A2ALL study, all data collection was performed through independently-funded efforts and often through chart review, which allowed for greater data granularity, completeness‚ and likely also validity. As LDLT continues to expand in the United States in the coming years, we hope that centers will strive for more consistent reporting of LD outcomes, as this is paramount to its continued success.

This study had several limitations. Although distinguishing the age 50–55 and age >55 groups is clinically relevant given that 55 y is the upper age limit for living liver donation at multiple U.S. programs, an age 50–55 age group was not practical in multivariable models because of low power. Nevertheless, our descriptive analyses help provide an overview of the landscape of older donor age in LDLT. An important aspect of LDLT safety is that of graft-to-recipient-weight ratio (GRWR) and of donor remnant volume. These parameters are not available in the UNOS database, yet are ubiquitously collected and factor heavily into the LD selection process. This prevented an evaluation of the potential interaction between donor age with graft and remnant volume, which would help better identify the circumstances in which increasing LD age may be more favorable, particularly given concerns raised in the study from Dayangac et al from Turkey.15 This should be further explored using multicenter data and also consider the influence of differences in donor selection practices in this context. For example, if GRWR were higher among LDLTs with older donors, then the effect of increasing donor age could be greater.

Beyond GRWR and donor remnant volume, other potentially relevant donor comorbidities and allograft characteristics were not available. For example, data on donor hypertension and diabetes are not collected for liver LDs in UNOS. Our study presents all donor comorbidities included in the database. Intraoperative factors such as warm ischemia time, portal venous pressures, surgical duration‚ and blood products used affect early posttransplant mortality in LDLT, but these are unfortunately not collected in UNOS.28 Moreover, biliary anatomy was not available, which would be relevant to the potential risk of biliary complications.29 Similarly, these factors may have also contributed to the differences in outcomes seen between LDLT and DDLT.30 For example, warm ischemia time may be longer due to LDLT being more technically difficult.28 Although these unmeasured donor and allograft factors have been shown to predict donor and recipient outcomes in LDLT, it is uncertain to what extent they differed according to donor age. If these were not important confounders, then inclusion in our multivariable models would not significantly alter the point estimates obtained for the effect of increasing donor age. Nevertheless, further evaluation of these more granular factors is warranted. However, given the marked center heterogeneity in LDLT practices, this can only be achieved through a large multicenter effort, which we hope our study will encourage. Despite its limitations, the UNOS database remains the only large multicenter LDLT data source in the United States. Lastly, the potential survival benefit of pursuing LDLT with an older LD over continuing to wait for DDLT must be investigated to better guide physician and patient decision-making.

In conclusion, despite the growth of LDLT in the United States, centers continue to be conservative in considering older living LDs, and our data support this approach. LD age ≥50 y remains an important independent predictor of graft failure and recipient mortality with no significant change in this association over time or according to other recipient factors. Importantly, the negative effects of increasing donor age on posttransplant outcomes after LDLT versus DDLT are not alike and it is possible that the benefits of LDLT over DDLT are lost in the setting of increasing LD age. From the recipient perspective, this decision to pursue LDLT with an older donor should be weighed against the risks of either continuing to wait for a DDLT or for a different LD option. Donor outcomes in this setting appear satisfactory though effects on long-term health are incompletely studied. Future multicenter studies are needed to establish the optimal donor-recipient pairing in the setting of marginal LDs.

REFERENCES

1. Lué A, Solanas E, Baptista P, et al. How important is donor age in liver transplantation? World J Gastroenterol. 2016;22:4966–4976.
2. Bittermann T, Goldberg DS. Quantifying the effect of transplanting older donor livers into younger recipients: the need for donor-recipient age matching. Transplantation. 2018;102:2033–2037.
3. Reese PP, Sonawane SB, Thomasson A, et al. Donor age and cold ischemia interact to produce inferior 90-day liver allograft survival. Transplantation. 2008;85:1737–1744.
4. Gao Q, Mulvihill MS, Scheuermann U, et al. Improvement in liver transplant outcomes from older donors: a US national analysis. Ann Surg. 2019;270:333–339.
5. Jiménez-Romero C, Caso Maestro O, Cambra Molero F, et al. Using old liver grafts for liver transplantation: where are the limits? World J Gastroenterol. 2014;20:10691–10702.
6. Renz JF, Kin C, Kinkhabwala M, et al. Utilization of extended donor criteria liver allografts maximizes donor use and patient access to liver transplantation. Ann Surg. 2005;242:556–63.
7. Health Resources & Services Administration. Is there an age limit for organ donation? 2021. Available at https://www.organdonor.gov/learn/who-can-donate/seniors. Accessed May 1, 2022.
8. Durand F, Levitsky J, Cauchy F, et al. Age and liver transplantation. J Hepatol. 2019;70:745–758.
9. Kubota T, Hata K, Sozu T, et al. Impact of donor age on recipient survival in adult-to-adult living-donor liver transplantation. Ann Surg. 2018;267:1126–1133.
10. Wang K, Jiang WT, Deng YL, et al. Effect of donor age on graft function and long-term survival of recipients undergoing living donor liver transplantation. Hepatobiliary Pancreat Dis Int. 2015;14:50–55.
11. Wei Q, Wang K, He Z, et al. Acute liver allograft rejection after living donor liver transplantation: risk factors and patient survival. Am J Med Sci. 2018;356:23–29.
12. Li J, Wang K. Application of elderly donor for liver transplantation. Int J Clin Exp Med. 2015;8:10315–10320.
13. Li C, Wen TF, Yan LN, et al. Safety of living donor liver transplantation using older donors. J Surg Res. 2012;178:982–987.
14. Goldaracena N, Sapisochin G, Spetzler V, et al. Live donor liver transplantation with older (≥50 years) versus younger (<50 years) donors: does age matter? Ann Surg. 2016;263:979–985.
15. Dayangac M, Taner CB, Yaprak O, et al. Utilization of elderly donors in living donor liver transplantation: when more is less? Liver Transpl. 2011;17:548–555.
16. Shah SA, Cattral MS, McGilvray ID, et al. Selective use of older adults in right lobe living donor liver transplantation. Am J Transplant. 2007;7:142–150.
17. Ghobrial RM, Freise CE, Trotter JF, et al. A2ALL Study Group. Donor morbidity after living donation for liver transplantation. Gastroenterology. 2008;135:468–476.
18. Pomposelli JJ, Goodrich NP, Emond JC, et al. Patterns of early allograft dysfunction in adult live donor liver transplantation: the A2ALL experience. Transplantation. 2016;100:1490–1499.
19. Olthoff KM, Abecassis MM, Emond JC, et al. Adult-to-Adult Living Donor Liver Transplantation Cohort Study Group. Outcomes of adult living donor liver transplantation: comparison of the Adult-to-Adult Living Donor Liver Transplantation Cohort Study and the national experience. Liver Transpl. 2011;17:789–797.
20. Olthoff KM, Smith AR, Abecassis M, et al. Defining long-term outcomes with living donor liver transplantation in North America. Ann Surg. 2015;262:465–75.
21. Levey AS, Coresh J, Greene T, et al. Chronic Kidney Disease Epidemiology Collaboration. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–254.
22. Kadohisa M, Inomata Y, Uto K, et al. Impact of donor age on the outcome of living-donor liver transplantation: special consideration to the feasibility of using elderly donors. Transplantation. 2021;105:328–337.
23. Han JH, Kim DG, Na GH, et al. Effect of donor-recipient age matching in living donor liver transplantation. Transplant Proc. 2015;47:718–722.
24. Ishiko T, Inomata Y, Beppu T, et al. Age and donor safety in living-donor liver transplant in 110 consecutive cases at 1 institute. Exp Clin Transplant. 2008;6:190–193.
25. Ahn J, Choi HJ, Park SE, et al. Noninvasive evaluations to estimate graft steatosis in living donor liver transplant for donor safety and successful outcome. Transplant Proc. 2022;54:374–379.
26. Hong YM, Yoon KT, Cho M, et al. Clinical usefulness of controlled attenuation parameter to screen hepatic steatosis for potential donor of living donor liver transplant. Eur J Gastroenterol Hepatol. 2017;29:805–810.
27. Qi Q, Weinstock AK, Chupetlovska K, et al. Magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) is a viable alternative to liver biopsy for steatosis quantification in living liver donor transplantation. Clin Transplant. 2021;35:e14339.
28. Tourky MS, Salman AA, Salman MA, et al. Intraoperative factors associated with early recipient death after adult-to-adult living donor liver transplant. Exp Clin Transplant. 2021;19:817–825.
29. Goldaracena N, Barbas AS. Living donor liver transplantation. Curr Opin Organ Transplant. 2019;24:131–137.
30. Tran L, Humar A. Current status of adult liver transplantation: utilization of living donor versus deceased donor graft. Curr Opin Organ Transplant. 2021;26:133–138.

Supplemental Digital Content

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.