This cohort study compares the survival outcomes associated with living-donor liver transplant vs deceased-donor liver transplant for transplant candidates with hepatocellular carcinoma.
Key Points
Question
Can the intention-to-treat survival benefit of a potential live donor be evaluated in a patient with hepatocellular carcinoma (HCC) who is on the waiting list for a liver transplant?
Findings
In this cohort study of nearly 4000 patients with HCC who were on a waiting list at transplant centers in Europe, Asia, the US, and Canada, living-donor liver transplant (LDLT) was an independent protective factor, reducing the risk of intention-to-treat death in 4 different settings. When LDLT was incorporated in the mathematical models, their discriminatory ability further improved.
Meaning
Findings from this study suggest that LDLT could potentially decrease the risk of death, by 33% to 49%, for patients with HCC awaiting a liver transplant.
Abstract
Importance
Living-donor liver transplant (LDLT) offers advantages over deceased-donor liver transplant (DDLT) of improved intention-to-treat outcomes and management of the shortage of deceased-donor allografts. However, conflicting data still exist on the outcomes of LDLT in patients with hepatocellular carcinoma (HCC).
Objective
To investigate the potential survival benefit of an LDLT in patients with HCC from the time of waiting list inscription.
Design, Setting, and Participants
This multicenter cohort study with an intention-to-treat design analyzed the data of patients aged 18 years or older who had an HCC diagnosis and were on a waiting list for a first transplant. Patients from 12 collaborative centers in Europe, Asia, and the US who were on a transplant waiting list between January 1, 2000, and December 31, 2017, composed the international cohort. The Toronto cohort comprised patients from 1 transplant center in Toronto, Ontario, Canada who were on a waiting list between January 1, 2000, and December 31, 2015. The international cohort centers performed either an LDLT or a DDLT, whereas the Toronto cohort center was selected for its capability to perform both LDLT and DDLT. The benefit of LDLT was tested in the 2 cohorts before and after undergoing an inverse probability of treatment weighting (IPTW) analysis. Data were analyzed from February 1 to May 31, 2020.
Main Outcomes and Measures
Intention-to-treat death was defined as a patient death that occurred for any reason and was calculated from the time of waiting list inscription for liver transplant to the last follow-up date (December 31, 2019). Four multivariable Cox proportional hazards regression models for intention-to-treat death were created.
Results
A total of 3052 patients were analyzed in the international cohort, of whom 2447 were men (80.2%) and the median (IQR) age at first referral was 58 (53-63) years. The Toronto cohort comprised 906 patients, of whom 743 were men (82.0%) and the median (IQR) age at first referral was 59 (53-63) years. In all the settings, LDLT was an independent protective factor, reducing the risk of overall death by 49% in the pre-IPTW analysis for the international cohort (HR, 0.51; 95% CI, 0.36-0.71; P < .001), 33% in the post-IPTW analysis for the international cohort (HR, 0.67; 95% CI, 0.53-0.85; P = .001), 43% in the pre-IPTW analysis for the Toronto cohort (HR, 0.57; 95% CI, 0.45-0.73; P < .001), and 48% in the post-IPTW analysis for the Toronto cohort (HR, 0.52; 95% CI, 0.42 to 0.65; P < .001). The discriminatory ability of the mathematical models further improved in all of the cases in which LDLT was incorporated.
Conclusions and Relevance
This study suggests that having a potential live donor could decrease the intention-to-treat risk of death in patients with HCC who are on a waiting list for a liver transplant. This benefit is associated with the elimination of the dropout risk and has been reported in centers in which both LDLT and DDLT options are equally available.
Introduction
Living-donor liver transplant (LDLT) offers advantages over deceased-donor liver transplant (DDLT) of improving intention-to-treat outcomes and managing the shortage of deceased-donor allografts.1,2 Nevertheless, the debate continues regarding the oncological benefit of an LDLT compared with a DDLT in patients with hepatocellular carcinoma (HCC).
Initial experiences have revealed higher recurrence rates in patients who underwent an LDLT compared with those who underwent a DDLT.3,4,5,6 Several reasons for these differences have been put forward. First, the fast-track approach should limit a suitable evaluation of the biological qualities of the tumor. Second, the parenchymal regeneration of the partial graft should favor tumor growth and spread. Third, the residual tumor should be present because of inferior vena cava preservation.7,8,9
However, later case series from Eastern and Western experiences have shown similar, or even better, results after an LDLT.10,11,12,13 Given these findings, one could hypothesize that patients with HCC who are on a waiting list are likely to obtain an intention-to-treat benefit from an LDLT.
This cohort study aimed to investigate the potential survival benefit of an LDLT in patients with HCC from the time of waiting list inscription. The benefit of live donation was tested in an international cohort; in a single-center cohort in Toronto, Canada; and in the same 2 populations after an inverse probability of treatment weighting (IPTW) analysis.
Methods
The present study was approved by the local ethics board of Sapienza University of Rome, Italy. Informed consent for the management of personal data was obtained for all patients at the time of the transplant; however, no consent was obtained for this study because of its retrospective nature. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Study Design, Setting, and Population
In this multicenter, retrospective cohort study with an intention-to-treat design, we investigated the data of adult patients (aged ≥18 years) with an HCC diagnosis who were on a waiting list for a first transplant. The potential survival benefit of an LDLT, calculated from the time of waiting list inscription, was explored in 4 different settings: (1) an international cohort representing 12 collaborative transplant centers in Europe, Asia, and the US; (2) the same international cohort but recalibrated after an IPTW analysis; (3) a Toronto cohort representing 1 transplant center; and (4) the same Toronto cohort but recalibrated after an IPTW analysis.
All participating centers in both cohorts were involved in the transplantation and transplant oncology fields and were part of The West-East Collaborative Liver Transplant Study Group. Specifically, the 12 centers that made up the international cohort were located in Europe (Université Catholique de Louvain, Brussels, Belgium; Sapienza University of Rome, Rome, Italy; University of Padua, Padua, Italy; University of Innsbruck, Innsbruck, Austria; and Universitätsmedizin Mainz, Mainz, Germany), Asia (Medanta-The Medicity, Delhi, India; University of Hong Kong, Hong Kong, China; Kyoto University, Kyoto, Japan; Chang Gung Memorial Hospital, Kaohsiung City, Taiwan; Kyushu University, Fukuoka, Japan; and Shulan Health Zhejiang University Hospital, Hangzhou, China), or the US (Columbia University, New York, New York). These 12 centers were dedicated to performing either an LDLT or a DDLT (eFigure in the Supplement).
The 1 center in the Toronto cohort was the University of Toronto in Toronto, Ontario, Canada. This center was selected as a separate cohort because of its capability to perform both complementary LDLT and DDLT. The international cohort was composed of 3052 patients with HCC who were on a waiting list for a first transplant between January 1, 2000, and December 31, 2017. The Toronto cohort included patients with HCC who were on a waiting list for a first transplant between January 1, 2000, and December 31, 2015. Excluded patients were those undergoing a retransplant and those with a graft treated with machine perfusion, with a donor after cardiac death, with mixed hepatocellular-cholangiocellular cancer, and with cholangiocellular cancer misdiagnosed as HCC.
Outcomes, Data Collection, and Definitions
The main outcome of the study was patient death that occurred for any reason from the time of waiting list inscription to the last follow-up. The last follow-up date was December 31, 2019.
Data were retrospectively retrieved from databases that are prospectively maintained by the transplant centers composing The West-East Collaborative Liver Transplant Study Group. One of us served as the guarantor of the data quality (Q.L., the data manager of the study group). Data errors and missingness across the databases were identified and solved, when possible, with specific queries to the centers. The missing data present in the evaluated databases are reported in eTable 1 in the Supplement.
All patients with a potential live donor during the waiting list period were included in the LDLT group. A potential live donor was defined as an individual who was suitable for donation after the screening phase and was enrolled for an imaging assessment.13 To preserve the intention-to-treat design of the study, we considered including in the LDLT group those patients with a suitable live donor and who eventually had a DDLT. In the LDLT category, we included (1) patients with a deceased donor available before the scheduled live donation date and (2) expected insufficient liver volume after the imaging assessment. In the DDLT category, we included patients with potential live donors who were considered unsuitable for donation before an imaging assessment.
The tumor upper limit for waiting list inscription differed among the participating centers. From 2000 onward, the initially applied restrictive Milan criteria were progressively abandoned. Overall, a more conservative approach was reported in DDLT, for which the University of California San Francisco and the Up-to-Seven restriction criteria were primarily used for listing and staying on the list. Often, in LDLT, no morphological limitations were adopted, except for the presence of major vascular invasion and extrahepatic disease localization (eTable 2 in the Supplement). In accordance with previous studies, we defined a high-volume center as one with a cutoff of 70 liver transplants per year.14
Statistical Analysis
Continuous variables were reported as medians (interquartile ranges [IQRs]), and categorical variables were described as numbers (percentages). Comparisons between groups were made using the Fisher exact test or χ2 test for categorical variables, as appropriate. The Mann-Whitney test was used for continuous variables. Missing data on study covariates (eTable 1 in the Supplement) always involved less than 10% of patients. In all cases, missing data were handled with a single imputation method. Specifically, a median of nearby points imputation was adopted. The median, instead of the mean, was adopted because of the skewed distribution of the managed variables.15
To compensate for the nonrandomized design of this study, we balanced (or corrected for potential confounders) the populations of the international cohort and Toronto cohort using the IPTW analysis. To compare the LDLT group with the DDLT group, we expressed continuous data as means (SDs) and based categorical data on the frequency distribution. The IPTW values were calculated using generalized boosted models as described by McCaffrey et al.16 This method is a machine learning technique that uses flexible estimation to adjust for a large number of covariates. Thirteen potential confounders were included in the boosted models: center case volume greater than 70 liver transplants per year, transplant year (before 2010), age, male sex, hepatitis C virus positive test result, hepatitis B virus positive test result, alcohol-related cirrhosis, waiting time duration, model for end-stage liver disease (MELD) score, diameter of the major lesion, number of HCC lesions, log10 α-fetoprotein level, and type of locoregional therapy.
To reduce the artificial increase of the sample size, and therefore the type I error rate (ie, increased number of false-positives) associated with the inflated sample size in the pseudo data, we used stabilized weights (SW) according to the following formula:
| SW = p/PS for the Study Group, and SW = (1 − p)/(1 − PS) for the Control Group, |
where p is the probability of cause without considering covariates, and PS is the propensity score.
Because P values can be biased from population size, results from the comparisons between covariate subgroups were reported as effect size (Cohen d value). The Cohen d values that were lower than 0.1 indicated very small differences between means, values between 0.1 and 0.3 indicated small differences, values between 0.3 and 0.5 indicated moderate differences, and values greater than 0.5 indicated large differences.17
Multivariable Cox proportional hazards regression models were run for the international cohort and Toronto cohort before and after the IPTW analysis to identify the risk factors for death evaluated from waiting list inscription. According to previous research on the same argument,13 the investigated variables were selected using a full model approach. In the pre-IPTW and post-IPTW analyses for the international cohorts, given the risk of data separation within the 12 liver transplant centers, a Cox proportional hazards regression model with mixed effects was created, in which the center was incorporated in the model as a cluster-specific random-effect variable.18 Hazard ratios (HRs) and 95% CIs were reported for significant variables.
The Akaike information criteria (AIC) value was calculated for the different regression models. The lowest AIC value was associated with the best discriminatory ability for the given model.19 Survival analyses were performed using the Kaplan-Meier method, and the log-rank test was adopted to compare the obtained survivals.
Variables with a 2-sided P < .05 were considered to be statistically significant. Statistical analyses were performed with SPSS, version 24.0 (IBM); Stata, version 14.0 (StataCorp LLC); and R.app 4.0.0 GUI 1.71 (R Foundation for Statistical Computing). Data were analyzed from February 1 to May 31, 2020.
Results
International Cohort Characteristics
Patient- and tumor-related characteristics of the international cohort population are summarized in Table 1. The median (IQR) follow-up for the entire series was 3.3 (1.5-6.6) years.
Table 1. Characteristics of Patients in the International and Toronto Cohorts .
| Variable | No. (%) | |||||
|---|---|---|---|---|---|---|
| International cohort | Toronto cohort | |||||
| Total | DDLT | LDLT | Total | DDLT | LDLT | |
| No. (%) | 3052 (100.0) | 2041 (66.9) | 1011 (33.1) | 906 (100.0) | 661 (73.0) | 245 (27.0) |
| Center volume >70 LT cases/y | 2227 (73.0) | 1587 (77.8) | 640 (63.3) | NA | NA | NA |
| LT performed before 2010 | 1458 (47.8) | 1102 (54.0) | 356 (35.2) | 487 (53.8) | 366 (55.4) | 121 (49.4) |
| Age at first referral, median (IQR), y | 58 (53-63) | 59 (54-63) | 57 (52-62) | 59 (53-63) | 59 (53-63) | 59 (54-63) |
| Male sex | 2447 (80.2) | 1707 (83.6) | 740 (73.2) | 743 (82.0) | 544 (82.3) | 199 (81.2) |
| Female sex | 605 (19.8) | 334 (16.4) | 271 (26.8) | 163 (18.0) | 117 (17.7) | 46 (18.8) |
| Underlying liver disease | ||||||
| HCV | 1383 (45.3) | 889 (43.6) | 494 (48.9) | 471 (52.0) | 335 (50.7) | 136 (55.5) |
| HBV | 880 (28.8) | 493 (24.2) | 387 (38.3) | 194 (21.4) | 169 (25.6) | 25 (10.2) |
| Alcohol-related cirrhosis | 630 (20.6) | 539 (26.4) | 91 (9.0) | 121 (13.4) | 85 (12.9) | 36 (14.7) |
| NASH | 204 (6.7) | 140 (6.9) | 64 (6.3) | 78 (8.6) | 49 (7.4) | 29 (11.8) |
| Other | 156 (5.1) | 129 (6.3) | 27 (2.7) | 45 (5.0) | 26 (3.9) | 19 (7.8) |
| Waiting time duration, median (IQR), mo | 4 (1-10) | 6 (3-13) | 1 (0-2) | 6 (3-11) | 6 (3-12) | 5 (3-8) |
| MELD score at first referral, median (IQR) | 12 (9-16) | 12 (9-16) | 12 (9-15) | 10 (8-14) | 10 (8-14) | 11 (8-14) |
| Tumor characteristics at first referral | ||||||
| Diameter of the major lesion, median (IQR), cm | 2.5 (1.8-3.7) | 2.5 (1.8-3.8) | 2.5 (1.7-3.6) | 2.7 (1.9-3.9) | 2.7 (1.9-3.7) | 2.9 (1.7-4.4) |
| No. of lesions, median (IQR) | 2 (1-3) | 1 (1-3) | 2 (1-3) | 1 (1-2) | 1 (1-2) | 1 (1-2) |
| Outside of MC | 959 (31.4) | 601 (29.4) | 358 (35.4) | 267 (29.5) | 185 (28.9) | 82 (33.5) |
| Tumor characteristics at LT or dropout | ||||||
| Diameter of the major lesion, median (IQR), cm | 2.2 (1.4-3.3) | 2.1 (1.2-3.0) | 2.5 (1.7-3.5) | 2.0 (0.0-3.6) | 2.0 (0.0-3.5) | 2.2 (0.0-3.9) |
| No. of lesions, median (IQR) | 2 (1-3) | 2 (1-3) | 2 (1-3) | 1 (0-3) | 1 (0-3) | 1 (0-2) |
| Outside of MC | 912 (29.9) | 553 (27.1) | 359 (35.5) | 269 (29.7) | 201 (30.4) | 68 (27.8) |
| AFP level, median (IQR), ng/mL | ||||||
| At first referral | 14 (5-60) | 11 (5-43) | 22 (7-114) | 11 (5-42) | 12 (5-43) | 10 (5-41) |
| At LT or dropout | 11 (4-56) | 10 (4-47) | 14 (4-77) | 11 (5-62) | 12 (5-62) | 10 (5-60) |
| LRT | 2369 (77.6) | 1819 (89.1) | 550 (54.4) | 623 (68.8) | 474 (71.7) | 149 (60.8) |
| Type of LRT | ||||||
| TACE | 1857 (60.8) | 1451 (71.1) | 406 (40.2) | 146 (16.1) | 104 (15.7) | 42 (17.1) |
| PEI | 486 (15.9) | 402 (19.7) | 84 (8.3) | 22 (2.4) | 18 (2.7) | 4 (1.6) |
| RFA | 743 (24.3) | 506 (24.8) | 237 (23.4) | 423 (46.7) | 329 (49.8) | 94 (38.4) |
| Hepatic resection | 266 (8.7) | 218 (10.7) | 48 (4.7) | 14 (1.5) | 14 (2.1) | 0 |
| Other | 251 (8.2) | 229 (11.2) | 22 (2.2) | 25 (2.8) | 14 (2.1) | 11 (4.5) |
| Dropout | 295 (7.8) | 295 (14.5) | 0 | 247 (27.3) | 213 (32.2) | 34 (13.9) |
| Death during waiting time | 159 (5.2) | 159 (7.8) | 0 | 70 (7.7) | 59 (8.9) | 11 (4.5) |
| Tumor progression | 80 (2.6) | 80 (3.9) | 0 | 129 (14.3) | 109 (16.5) | 20 (8.2) |
| Posttransplant recurrence | 360 (11.8) | 223 (12.9) | 137 (13.6) | 116 (17.6) | 88 (19.7) | 28 (13.3) |
Abbreviations: AFP, α-fetoprotein; DDLT, deceased-donor liver transplant; HBV, hepatitis B virus; HCV, hepatitis C virus; IQR, interquartile range; LDLT, living-donor liver transplant; LRT, locoregional therapy; LT, liver transplant; MC, Milan criteria; MELD, model for end-stage liver disease; NA, not applicable; NASH, nonalcoholic steatohepatitis; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; TACE, transarterial chemoembolization.
Among the 3052 patients in the international cohort, 1011 (33.1%) had a potential LDLT and 2041 (66.9%) had a potential DDLT. The median (IQR) age at first referral was 58 (53-63) years, and the group was composed of 2447 men (80.2%) and 605 women (19.8%). Exclusive DDLT or LDLT activity was reported in 4 centers, and mixed activity was observed in 4 other centers. However, the DDLT activity was prominent (>90%) in 3 centers, whereas the Hong Kong center was the only 1 with a real mixed LDLT and DDLT activity (54.5% vs 45.5%) (eFigure in the Supplement).
Several differences existed between the LDLT and DDLT groups, such as transplant candidate features, underlying liver pathology, tumor characteristics, and management during the waiting period. At first referral, patients in both the DDLT and LDLT groups were outside of the Milan criteria (601 [29.4%] vs 358 [35.4%]; P < .001). As expected, the median (IQR) waiting time was shorter in the LDLT vs DDLT group (1 [0-2] vs 6 [3-13] months; P < .001).
The number of dropouts was markedly superior in the DDLT group vs the LDLT group (295 vs 0 patients; P < .001). Figure 1 displays the reasons for the dropouts. A total of 295 patients (14.5%) dropped out from the DDLT group as a result of death while waiting for a transplant (159 [53.9%]), tumor progression (80 [27.1%]), or other factors (56 [19.0%]). Interestingly, no patients dropped out because of tumor progression or death while on the waiting list in the LDLT group. Twenty-seven candidates for LDLT (2.7%) moved to the DDLT waiting list because of the availability of a deceased donor (3 [0.3%]) or the impossibility of receiving a live donation because of insufficient liver volume (24 [2.4%]).
Figure 1. Flowchart of Patients in the International and Toronto Cohorts.
DDLT indicates deceased-donor liver transplant; HCC, hepatocellular carcinoma; LDLT, living-donor liver transplant; LT, liver transplant.
Toronto Cohort Characteristics
Patient- and tumor-related characteristics of the Toronto cohort population are summarized in Table 1. The median (IQR) follow-up for the entire series was 3.2 (1.4-7.5) years.
Among the 906 patients considered in the Toronto cohort, 245 (27.0%) had a potential LDLT and 661 (73.0%) had a potential DDLT. The median (IQR) age at first referral was 59 (53-63) years, and the group was composed of 743 men (82.0%) and 163 women (18.0%). Several differences existed between the 2 groups. The median (IQR) waiting time was shorter in the LDLT than in the DDLT group (5 [3-8] vs 6 [3-12] months; P = .006). The percentage of dropouts was superior in the DDLT group vs the LDLT group (32.2% vs 13.9%; P < .001). In Figure 1, the reasons for the dropouts are detailed.
In the DDLT group of the Toronto cohort, 213 patients (32.2%) dropped out as a result of death while waiting for a transplant (59 [27.7%]), tumor progression (109 [51.2%]), or other factors (45 [21.1%]). In the LDLT group, 34 patients (13.9%) dropped out as a result of tumor progression (20 [58.8%]), death while waiting for a transplant (11 [32.4%]), or other factors (3 [8.8%]). Seventy-three LDLT candidates (8.1%) moved to the DDLT program because of the availability of a deceased donor before the scheduled date of LDLT.
IPTW Analysis
To minimize the selection biases caused by the nonrandomized design of this cohort study, the international and Toronto cohorts were artificially balanced using the IPTW method. As reported in Table 2, both of the cohorts were balanced. Specifically, all the variables tested in the post-IPTW analysis of the international cohort showed only small differences, with 7 variables showing very small differences. In the post-IPTW analysis of the Toronto cohort, all the variables except for 1 showed very small differences.
Table 2. Association of Inverse Probability of Treatment Weighting (IPTW) With the Variables Used for Balancing the International and Toronto Cohorts.
| Variable | Pre-IPTW, No. (%) | Post-IPTW, No. (%) | ||||
|---|---|---|---|---|---|---|
| DDLT | LDLT | Cohen da | DDLT | LDLT | Cohen da | |
| International cohort | ||||||
| No. of patients | 2041 | 1011 | NA | 1748 | 741 | NA |
| Center volume >70 LT/y | 1587 (77.8) | 640 (63.3) | –0.33 | 1329 (76.0) | 543 (73.3) | –0.06 |
| LT before 2010 | 1102 (54.0) | 356 (35.2) | –0.38 | 877 (50.2) | 344 (46.4) | –0.08 |
| Age, mean (SD), yb | 58 (8) | 56 (8) | –0.25 | 57 (8) | 57 (7) | –0.07 |
| Male sex | 1707 (83.6) | 740 (73.2) | –0.26 | 1458 (83.4) | 576 (77.7) | –0.15 |
| Female sex | 334 (16.4) | 271 (26.8) | –0.26 | 290 (16.6) | 165 (22.3) | –0.15 |
| HCV | 889 (43.6) | 494 (48.9) | –0.11 | 760 (43.5) | 400 (54.0) | 0.21 |
| HBV | 493 (24.2) | 387 (38.3) | 0.31 | 475 (27.2) | 223 (30.1) | 0.07 |
| Alcohol-related cirrhosis | 539 (26.4) | 91 (9.0) | –0.43 | 406 (23.2) | 111 (15.0) | –0.21 |
| Waiting time duration, mean (SD), mo | 13 (23) | 2 (5) | –0.53 | 11 (21) | 6 (10) | –0.26 |
| MELD score, mean (SD)b | 13 (7) | 13 (5) | –0.10 | 13 (6) | 13 (5) | –0.03 |
| Diameter of major lesion, mean (SD), cmb | 3.1 (2.1) | 3.1 (2.3) | 0.01 | 3.0 (2.0) | 2.9 (2.1) | –0.04 |
| No. of lesions, mean (SD)b | 2 (5) | 3 (7) | 0.09 | 2 (5) | 2 (5) | 0.00 |
| AFP level, mean (SD), ng/mLb | 354 (4850) | 828 (7445) | 0.39 | 384 (5267) | 474 (5463) | 0.20 |
| LRT | 1819 (89.1) | 550 (54.4) | –0.83 | 1477 (84.5) | 541 (72.9) | –0.27 |
| Toronto cohort | ||||||
| No. of patients | 661 | 245 | NA | 690 | 317 | NA |
| Center volume >70 LT/y | NA | NA | NA | NA | NA | NA |
| LT before 2010 | 366 (55.4) | 121 (49.4) | –0.12 | 361 (52.3) | 179 (56.3) | –0.03 |
| Age, mean (SD), yb | 58 (8) | 58 (8) | –0.25 | 58 (8) | 57 (8) | 0.02 |
| Male sex | 544 (82.3) | 199 (81.2) | –0.03 | 567 (82.2) | 261 (82.3) | –0.01 |
| Female sex | 117 (17.7) | 46 (18.8) | –0.03 | 123 (17.8) | 56 (17.7) | –0.01 |
| HCV | 335 (50.7) | 136 (55.5) | 0.10 | 348 (50.4) | 175 (55.0) | 0.07 |
| HBV | 169 (25.6) | 25 (10.2) | –0.37 | 177 (25.7) | 29 (9.1) | –0.21 |
| Alcohol-related cirrhosis | 85 (12.9) | 36 (14.7) | 0.05 | 91 (13.2) | 51 (16.1) | –0.01 |
| Waiting time duration, mean (SD), mo | 8 (7) | 7 (6) | –0.53 | 8 (7) | 7 (7) | –0.01 |
| MELD score, mean (SD)b | 11 (5) | 12 (5) | –0.10 | 11 (5) | 12 (5) | 0.04 |
| Diameter major lesion, mean (SD), cmb | 3.0 (1.8) | 3.5 (2.6) | 0.01 | 3.0 (1.8) | 3.6 (2.7) | 0.06 |
| No. of lesions, mean (SD)b | 2 (1) | 2 (2) | 0.09 | 2 (1) | 2 (2) | –0.01 |
| AFP level, mean (SD), ng/mLb | 386 (6097) | 222 (940) | 0.39 | 385 (6137) | 232 (1002) | –0.05 |
| LRT | 474 (71.7) | 149 (60.8) | –0.24 | 502 (72.6) | 184 (57.9) | –0.10 |
Abbreviations: AFP, α-fetoprotein; HCV, hepatitis C virus; HBV, hepatitis B virus; LRT, locoregional therapy; LT, liver transplant; MELD, model for end-stage liver disease; NA, not applicable.
Cohen d values that were lower than 0.1 indicated very small differences between means; between |0.1| and |0.3|, small differences; between 0.3 and 0.5, moderate differences; and greater than 0.5, large differences.
At first referral.
LDLT and Multivariable Models
Four multivariable Cox proportional hazards regression models for the risk of intention-to-treat death were constructed using the pre-IPTW and post-IPTW analyses for the international cohort and the Toronto cohort (Table 3). In the pre-IPTW international cohort, a live donor availability was a protective factor for intention-to-treat death (HR, 0.51; 95% CI, 0.36-0.71; P < .001). Similar results were reported after the IPTW analysis, with the confirmed protective factor of LDLT (HR, 0.67; 95% CI, 0.53-0.85; P = .001). When the Toronto cohort data were analyzed, the protective factor of live donor availability was confirmed again (pre-IPTW: HR, 0.57; [95% CI, 0.45-0.73; P < .001]; post-IPTW: HR, 0.52 [95% CI, 0.42 to 0.65; P < .001]).
Table 3. Multivariable Models for the Risk of Patient Death From the Time of Waiting List Inscription.
| Variable | Pre-IPTW International cohorta | Post-IPTW International cohorta | Pre-IPTW Toronto cohort | Post-IPTW Toronto cohort | ||||
|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| LDLT | 0.51 (0.36 to 0.71) | <.001 | 0.67 (0.53 to 0.85) | .001 | 0.57 (0.45 to 0.73) | <.001 | 0.52 (0.42 to 0.65) | <.001 |
| Diameter of major lesion, cm | 1.07 (1.03 to 1.12) | <.001 | 1.03 (0.99 to 1.07) | .20 | 1.08 (1.03 to 1.12) | .001 | 1.07 (1.03 to 1.11) | .001 |
| Log10 AFP level | 1.21 (1.13 to 1.28) | <.001 | 1.21 (1.11 to 1.33) | <.001 | 1.33 (1.18 to 1.49) | <.001 | 1.31 (1.18 to 1.46) | <.001 |
| MELD score | 1.02 (1.00 to 1.03) | .02 | 1.03 (1.02 to 1.04) | <.001 | 1.03 (1.01 to 1.05) | .01 | 1.03 (1.00 to 1.05) | .02 |
| Patient age, y | 1.01 (1.00 to 1.03) | .02 | 1.01 (1.00 to 1.02) | .03 | 1.02 (1.01 to 1.03) | .003 | 1.02 (1.01 to 1.03) | .002 |
| Waiting time duration, mo | 0.99 (0.99 to 1.00) | .05 | 1.11 (1.08 to 1.14) | <.001 | 0.99 (0.98 to 1.01) | .21 | 0.99 (0.98 to 1.00) | .21 |
| No. of lesions | 1.00 (.999 to 1.01) | .07 | 1.00 (0.99 to 1.02) | .91 | 0.96 (0.90 to 1.03) | .23 | 0.96 (0.90 to 1.02) | .16 |
| HBV-related cirrhosis | 0.82 (0.64 to 1.06) | .14 | 0.87 (0.70 to 1.08) | .21 | 0.69 (0.49 to 0.99) | .04 | 0.67 (0.48 to 0.94) | .02 |
| HCV-related cirrhosis | 1.15 (0.93 to 1.42) | .19 | 0.91 (0.75 to 1.10) | .34 | 1.17 (0.87 to 1.59) | .30 | 1.15 (0.86 to 1.53) | .34 |
| Male sex | 0.95 (0.84 to 1.08) | .45 | 1.07 (0.89 to 1.27) | .49 | 0.91 (0.72 to 1.16) | .46 | 0.92 (0.73 to 1.15) | .45 |
| Female sex | 1.05 (0.93 to 1.19) | .45 | 0.93 (0.79 to 1.12) | .49 | 1.10 (0.86 to 1.39) | .46 | 1.09 (0.87 to 1.37) | .45 |
| Alcohol-related cirrhosis | 1.13 (0.80 to 1.58) | .49 | 1.02 (0.84 to 1.25) | .83 | 1.00 (0.68 to 1.46) | .99 | 0.91 (0.64 to 1.31) | .62 |
| Period of LT before 2010 | 0.92 (0.69 to 1.21) | .54 | 0.89 (0.76 to 1.03) | .11 | 1.05 (0.85 to 1.30) | .65 | 1.07 (0.87 to 1.31) | .54 |
| LRT | 1.09 (0.81 to 1.48) | .55 | 1.00 (0.81 to 1.22) | .96 | 0.79 (0.64 to 0.97) | .03 | 0.80 (0.66 to 0.98) | .03 |
| Center volume >70 LT/y | 1.10 (0.77 to 1.56) | .61 | 1.00 (0.85 to 1.18) | .99 | NA | NA | NA | NA |
Abbreviations: AFP, α-fetoprotein; HBV, hepatitis B virus; HR, hazard ratio; IPTW, inverse probability of treatment weighting; LDLT, living-donor liver transplant; LRT, locoregional therapy; LT, liver transplant; MELD, model for end-stage liver disease; NA, not applicable; NASH, nonalcoholic steatohepatitis.
Cox proportional hazards regression model with mixed effects (cluster-specific random-effects variable = LT center) was used.
According to the HRs observed in all 4 models, a living donation was associated with a 33% to 49% reduction of the intention-to-treat death risk. When the AIC value was calculated for each model, the introduction of the variable live donation always decreased the AIC value in the pre-IPTW analysis (13 204.6 without LDLT to 13 124.8 with LDLT) and post-IPTW analysis (13 262.7 without LDLT to 13 202.7 with LDLT) for the international cohort as well as the pre-IPTW (5579.5 without LDLT to 5490.3 with LDLT) and post-IPTW analyses (6299.2 without LDLT to 6259.9 with LDLT) for the Toronto cohort. These AIC values show an increase in the discriminative ability of the models.
Intention-to-Treat Survival Rates
During the follow-up period, 938 of 3052 patients (30.7%) in the international cohort and 447 of 906 patients (49.3%) in the Toronto cohort died. Specifically, 246 patients in the international cohort died during the waiting period or after dropout (184 HCC-related vs 62 nontumor-related deaths), and 692 died after transplant (242 HCC-related vs 450 nontumor-related deaths). In the Toronto cohort, 241 patients died during the waiting period or after dropout (129 HCC-related vs 112 nontumor-related deaths), and 206 died after transplant (96 HCC-related vs 110 nontumor-related deaths).
In the pre-IPTW analysis for the international cohort, the 5-year intention-to-treat survival rate was 79% and the 10-year rate was 65% for the LDLT group (P < .001), whereas the 5-year rate was 70% and the 10-year rate was 52% for the DDLT group (P < .001) (Figure 2). When the survival rates were recalculated in the post-IPTW population, only limited changes were observed. In the post-IPTW international cohort population, the 5-year intention-to-treat survival rate was 79% and the 10-year rate was 64% for the LDLT group (P < .001), whereas the 5-year rate was 72% and the 10-year rate was 49% for the DDLT group (P < .001).
Figure 2. Intention-to-Treat Patient Survival Rates Before and After Inverse Probability of Treatment Weighting (IPTW) in the International Cohort (A and B) and Toronto Cohort (C and D).
DDLT indicates deceased-donor liver transplant; LDLT, living-donor liver transplant.
In the pre-IPTW analysis for the Toronto cohort, the 5-year intention-to-treat survival rate was 62% and the 10-year rate was 54% for the LDLT group (P = .04), whereas the 5-year rate was 55% and the 10-year rate was 42% for the DDLT group (P = .04) (Figure 2). In the post-IPTW population, the 5-year intention-to-treat survival rate was 66% and the 10-year rate was 52% for the LDLT group (P < .001), whereas the 5-year rate was 57% and the 10-year rate was 41% for the DDLT group (P < .001).
When only the 959 patients initially outside of the Milan criteria were considered in the international cohort, the 5-year intention-to-treat survival rate was 71% and the 10-year rate was 62% for the LDLT group (P < .001), whereas the 5-year rate was 65% and the 10-year rate was 49% in the DDLT group (P < .001). In the 267 patients outside of the Milan criteria in the Toronto cohort, the 5-year intention-to-treat survival rate was 61% and the 10-year rate was 57% for the LDLT group (P = .44), whereas the 5-year rate was 48% and the 10-year rate was 40% for the DDLT group (P = .44).
Posttransplant Recurrence Rates
During the posttransplant follow-up period, 360 of 2733 patients (13.2%) in the international cohort and 116 of 659 patients (17.6%) in the Toronto cohort experienced HCC recurrence. In the pre-IPTW analysis for the international cohort, the 5-year recurrence rate was 16% and the 10-year rate was 17% for the LDLT group (P = .57), whereas the 5-year rate was 17% and the 10-year rate was 21% for the DDLT group (P = .57). In the post-IPTW population, the 5-year recurrence rate was 13% and the 10-year rate was 18% for the LDLT group (P = .07), whereas the 5-year rate was 16% and the 10-year rate was 22% for the DDLT group (P = .07).
In the pre-IPTW analysis for the Toronto cohort, the 5-year recurrence rate was 13% and the 10-year rate was 21% for the LDLT group (P = .14), whereas the 5-year rate was 20% and the 10-year rate was 23% for the DDLT group (P = .14). In the post-IPTW analysis for the Toronto cohort, the 5-year recurrence rate was 13% and the 10-year rate was 22% for the LDLT group (P = .048), whereas the 5-year rate was 18% and the 10-year rate was 23% for the LDLT group (P = .048).
When only the 890 patients initially outside of the Milan criteria who underwent a transplant were considered in the international cohort, the 5-year recurrence rate was 27% and the 10-year rate was 26% for the LDLT group (P = .50), whereas the 5-year rate was 30% and the 10-year rate was 28% for the DDLT group (P = .50). In the 204 patients outside of the Milan criteria in the Toronto cohort, the 5-year recurrence rate was 23% and the 10-year rate was 32% for the LDLT group (P = .36), whereas the 5-year rate was 23% and the 10-year rate was 35% for the DDLT group (P = .36).
Discussion
In the present study, a potential living donation was beneficial to the intention-to-treat survival of candidates for liver transplant. Analyzing a large international population composed of centers in Europe, Asia, and the US revealed a 49% reduction in death risk from the time of waiting list inscription. After balancing the model for potential confounders using the IPTW method, the risk reduction remained high (ie, 33%). The 12 centers of the international cohort performed either a DDLT or an LDLT exclusively. When the protective factor of an LDLT was tested in the Toronto cohort, the risk reduction was 43% in the pre-IPTW population and 48% in the post-IPTW population. The Toronto center was chosen as a separate cohort because of its capability in performing both LDLT and DDLT as complementary approaches.
When the first living-donation case series were reported (mainly in the Eastern countries) 2 aspects of the procedure immediately emerged: the shorter, or even absent, waiting time and the concomitant decreased dropout risk of the transplant candidates.20 Patients with a living donor were not affected by graft scarcity and did not add any competitive harm to the list of patients who were waiting for a DDLT.21 One could even argue that a living donor gives 2 organs back to the organ pool. However, living donation also presented risks to patients in need of a transplant with poorly investigated biological qualities of the tumor (fast-track approach).9,22 Consequently, poor results were initially reported in LDLT in HCC case series. For example, the Adult-to-Adult Living Donor Liver Transplantation Cohort Study (with 58 LDLTs vs 34 DDLTs) showed a higher 5-year recurrence rate in an LDLT compared with a DDLT (29% vs 0%).3 A meta-analysis (with 633 LDLTs vs 1232 DDLTs) also showed an increased risk for HCC recurrence after an LDLT (HR, 1.59; 95% CI, 1.02-2.49).5
A study from Hong Kong (with 43 LDLTs vs 17 DDLTs) also reported disappointing results after live donation (3-year recurrence rates, 29% vs 0%).6 In this study, salvage transplant was a significant risk factor in recurrence.6 Another study that compared the Hong Kong with the Italian case series again found the unfavorable outcomes of salvage for HCC recurrence in LDLT.23
However, all of these results were potentially biased by the small number of cases and by more advanced tumors typically being transplanted in LDLT case series.4
Recent studies of more homogeneous LDLT and DDLT populations showed the benefit of live donation.10,11,12,13 A French multicenter study (with 79 LDLTs vs 782 DDLTs) showed no dropouts (0% vs 20.7%; P < .001), better intention-to-treat overall survival rates, and no added recurrence risk in patients who underwent LDLT.10 Another French study (with 36 LDLTs vs 147 DDLTs) found fewer dropouts and shorter median waiting times in patients who underwent LDLT.11 Using a propensity score match, a Hong Kong group reported similar recurrence-free survival rates, although more advanced tumors were transplanted in the LDLT group.12 The Toronto study (with 219 LDLTs vs 632 DDLTs) reported that having a potential live donor was protective for the risk of intention-to-treat death (HR, 0.67; 95% CI, 0.53-0.86).13
An international study based on 4089 patients recalibrating the Hazard Associated with Liver Transplantation for Hepatocellular Carcinoma database showed that undergoing a transplant in Asia reduced the risk of intention-to-treat death by 43%.24 All of the Asian patients in the report had an LDLT.24 A recent meta-analysis (with 5376 patients) found that LDLT was associated with better 5-year intention-to-treat patient survival rates (relative risk, 1.11; 95% CI, 1.01-1.22; P = .04), whereas the 5-year recurrence rates were similar (relative risk, 0.85; 95% CI, 0.56-1.31; P = .47).25
Results of the present study were consistent with all of these previous findings, reporting the beneficial outcome of having a potential live donor for transplant candidates with HCC. Compared with previous studies, all of which were based on relatively small case series, the present study examined the data of almost 4000 patients who were on a waiting list for a transplant; therefore, this study may be the largest cohort study on this topic.
This analysis was supported by the IPTW approach, which balanced the potential biases in the international cohort and Toronto cohort. The reason for such an outcome was associated with the absence of dropouts in the LDLT group. The number of patients who switched to the DDLT group was minimal. The centers of the international cohort were almost exclusively LDLT or DDLT centers. Therefore, only by merging the data were we able to evaluate the outcome of living donation in the setting of HCC.
The combination of different experiences, management, and patient- and tumor-related characteristics may compel one to question the validity of the observed results. To minimize these potential limitations, we created Cox proportional hazards regression models with mixed effects, in which the transplant center was incorporated in the model as a cluster-specific random-effect variable.
Moreover, live donation was further evaluated in a single center that performed both live- and deceased-donor procedures without a particular attitude about or preference for one over another procedure. An LDLT was associated with a marked reduction in the risk of intention-to-treat death (pre-IPTW: 43%; post-IPTW: 48%) in the Toronto cohort.
Divergent experiences all converged to a similar 40% to 50% reduction in intention-to-treat death risk. This result was also observed in areas in which live donation is less commonly performed, such as the Toronto cohort. Therefore, the benefits of LDLT should motivate the implementation of live donations in Asia.26,27,28
Another critical aspect reported in this study is that patients who had an LDLT did not show an increased recurrence risk. An LDLT did not have unfavorable intent-to-treat survival and recurrence rates for the subgroup of patients who were initially outside of the Milan criteria.
All of these data are relevant, given that several centers adopted more liberal criteria (ie, more advanced tumors) in the case of live donation availability. Such consideration further underlines the importance of a live donor program in transplant centers that manage a high volume of patients with a tumor.
An LDLT program must have high standards of safety for the donor to minimize any potential risk of morbidity.29 Therefore, the performance of an LDLT should be restricted to larger hepatopancreatobiliary and transplant programs, which have high transplant and hepatic surgery expertise.30 Opportunities for training liver surgeons to master both liver resection and transplant must be made available.
Limitations
This study has some limitations. First, the retrospective study design and the potential initial selection biases could result in missing confounding factors, potentially favoring the outcomes in patients undergoing LDLT. The IPTW method and the evaluation of survival benefits of LDLT in a single-center cohort, in which both LDLT and DDLT are equally performed, were used to minimize these potential adverse effects. Second, the real number of patients who died or experienced tumor progression before listing could not be ascertained because of the absence of an available live donor. This limitation was particularly true in centers that performed LDLT exclusively. Therefore, a certain underestimation of the intention-to-treat mortality should be contemplated in the LDLT group.
Third, several differences were observed between patients in Eastern and Western countries, including HCC pretransplant treatments, genetics, and underlying liver disease.31 To minimize a potential geographical outcome, we constructed a Cox proportional hazards regression model with mixed effects in which the population was hierarchically clustered according to the center in which the transplant was performed. Moreover, the benefit of the live donation was investigated in a homogeneous single center in which both the LDLT and DDLT options were available.
Conclusions
This cohort study showed that a potential live donor availability decreased the risk of intention-to-treat death by as much as 49% in patients with HCC who were on a waiting list for a liver transplant. This benefit was associated with the reduction or even elimination of dropout rates. These results were supported by an IPTW analysis and were reported in centers in which both LDLT and DDLT options were equally available. Transplant programs worldwide should be encouraged to expand their live donor programs to manage patients with HCC.
eTable 1. Missing Data Observed in the Databases Investigated
eTable 2. Upper Limit of Transplantability in the Different Centers
eFigure. Deceased- and Living-Donor Liver Transplantation Activity in the 12 Centers Composing the International Cohort
References
- 1.Shah SA, Levy GA, Greig PD, et al. Reduced mortality with right-lobe living donor compared to deceased-donor liver transplantation when analyzed from the time of listing. Am J Transplant. 2007;7(4):998-1002. doi: 10.1111/j.1600-6143.2006.01692.x [DOI] [PubMed] [Google Scholar]
- 2.Humar A, Ganesh S, Jorgensen D, et al. Adult living donor versus deceased donor liver transplant (LDLT versus DDLT) at a single center: time to change our paradigm for liver transplant. Ann Surg. 2019;270(3):444-451. doi: 10.1097/SLA.0000000000003463 [DOI] [PubMed] [Google Scholar]
- 3.Fisher RA, Kulik LM, Freise CE, et al. ; A2ALL Study Group . Hepatocellular carcinoma recurrence and death following living and deceased donor liver transplantation. Am J Transplant. 2007;7(6):1601-1608. doi: 10.1111/j.1600-6143.2007.01802.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kulik LM, Fisher RA, Rodrigo DR, et al. ; A2ALL Study Group . Outcomes of living and deceased donor liver transplant recipients with hepatocellular carcinoma: results of the A2ALL cohort. Am J Transplant. 2012;12(11):2997-3007. doi: 10.1111/j.1600-6143.2012.04272.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Grant RC, Sandhu L, Dixon PR, Greig PD, Grant DR, McGilvray ID. Living vs. deceased donor liver transplantation for hepatocellular carcinoma: a systematic review and meta-analysis. Clin Transplant. 2013;27(1):140-147. doi: 10.1111/ctr.12031 [DOI] [PubMed] [Google Scholar]
- 6.Lo CM, Fan ST, Liu CL, Chan SC, Ng IO, Wong J. Living donor versus deceased donor liver transplantation for early irresectable hepatocellular carcinoma. Br J Surg. 2007;94(1):78-86. doi: 10.1002/bjs.5528 [DOI] [PubMed] [Google Scholar]
- 7.Moon DB, Lee SG, Hwang S, et al. No-touch en bloc right lobe living-donor liver transplantation with inferior vena cava replacement for hepatocellular carcinoma close to retrohepatic inferior vena cava: case report. Transplant Proc. 2013;45(8):3135-3139. doi: 10.1016/j.transproceed.2013.08.052 [DOI] [PubMed] [Google Scholar]
- 8.Taketomi A, Shirabe K, Toshima T, et al. The long-term outcomes of patients with hepatocellular carcinoma after living donor liver transplantation: a comparison of right and left lobe grafts. Surg Today. 2012;42(6):559-564. doi: 10.1007/s00595-011-0086-4 [DOI] [PubMed] [Google Scholar]
- 9.Ng KK, Lo CM, Chan SC, Chok KS, Cheung TT, Fan ST. Liver transplantation for hepatocellular carcinoma: the Hong Kong experience. J Hepatobiliary Pancreat Sci. 2010;17(5):548-554. doi: 10.1007/s00534-009-0165-8 [DOI] [PubMed] [Google Scholar]
- 10.Azoulay D, Audureau E, Bhangui P, et al. Living or brain-dead donor liver transplantation for hepatocellular carcinoma: a multicenter, western, intent-to-treat cohort study. Ann Surg. 2017;266(6):1035-1044. doi: 10.1097/SLA.0000000000001986 [DOI] [PubMed] [Google Scholar]
- 11.Bhangui P, Vibert E, Majno P, et al. Intention-to-treat analysis of liver transplantation for hepatocellular carcinoma: living versus deceased donor transplantation. Hepatology. 2011;53(5):1570-1579. doi: 10.1002/hep.24231 [DOI] [PubMed] [Google Scholar]
- 12.Wong TCL, Ng KKC, Fung JYY, et al. Long-term survival outcome between living donor and deceased donor liver transplant for hepatocellular carcinoma: intention-to-treat and propensity score matching analyses. Ann Surg Oncol. 2019;26(5):1454-1462. doi: 10.1245/s10434-019-07206-0 [DOI] [PubMed] [Google Scholar]
- 13.Goldaracena N, Gorgen A, Doyle A, et al. Live donor liver transplantation for patients with hepatocellular carcinoma offers increased survival vs. deceased donation. J Hepatol. 2019;70(4):666-673. doi: 10.1016/j.jhep.2018.12.029 [DOI] [PubMed] [Google Scholar]
- 14.Avolio AW, Franco A, Schlegel A, et al. Development and validation of a comprehensive model to estimate early allograft failure among patients requiring early liver retransplant. JAMA Surg. 2020;155(12):e204095. doi: 10.1001/jamasurg.2020.4095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhang Z. Missing data imputation: focusing on single imputation. Ann Transl Med. 2016;4(1):9. doi: 10.21037/atm-20-3623 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.McCaffrey DF, Griffin BA, Almirall D, Slaughter ME, Ramchand R, Burgette LF. A tutorial on propensity score estimation for multiple treatments using generalized boosted models. Stat Med. 2013;32(19):3388-3414. doi: 10.1002/sim.5753 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Burnand B, Kernan WN, Feinstein AR. Indexes and boundaries for “quantitative significance” in statistical decisions. J Clin Epidemiol. 1990;43(12):1273-1284. doi: 10.1016/0895-4356(90)90093-5 [DOI] [PubMed] [Google Scholar]
- 18.Austin PC. A tutorial on multilevel survival analysis: methods, models and applications. Int Stat Rev. 2017;85(2):185-203. doi: 10.1111/insr.12214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Stone M. Akaike’s criteria. In: Armitage P, Colton T, eds. Encyclopedia of Biostatistics. Wiley; 1998:123-124. [Google Scholar]
- 20.Fisher RA. Living donor liver transplantation: eliminating the wait for death in end-stage liver disease? Nat Rev Gastroenterol Hepatol. 2017;14(6):373-382. doi: 10.1038/nrgastro.2017.2 [DOI] [PubMed] [Google Scholar]
- 21.Volk ML, Vijan S, Marrero JA. A novel model measuring the harm of transplanting hepatocellular carcinoma exceeding Milan criteria. Am J Transplant. 2008;8(4):839-846. doi: 10.1111/j.1600-6143.2007.02138.x [DOI] [PubMed] [Google Scholar]
- 22.Lee KW, Yi NJ, Suh KS. Section 5: further expanding the criteria for HCC in living donor liver transplantation: when not to transplant: SNUH experience. Transplantation. 2014;97(suppl 8):S20-S23. doi: 10.1097/01.tp.0000446269.20934.d3 [DOI] [PubMed] [Google Scholar]
- 23.Lai Q, Avolio AW, Lerut J, et al. Recurrence of hepatocellular cancer after liver transplantation: the role of primary resection and salvage transplantation in East and West. J Hepatol. 2012;57(5):974-979. doi: 10.1016/j.jhep.2012.06.033 [DOI] [PubMed] [Google Scholar]
- 24.Firl DJ, Sasaki K, Agopian VG, et al. ; European Hepatocellular Cancer Liver Transplant Study Group . Charting the path forward for risk prediction in liver transplant for hepatocellular carcinoma: international validation of HALTHCC among 4,089 patients. Hepatology. 2020;71(2):569-582. doi: 10.1002/hep.30838 [DOI] [PubMed] [Google Scholar]
- 25.Zhu B, Wang J, Li H, Chen X, Zeng Y. Living or deceased organ donors in liver transplantation for hepatocellular carcinoma: a systematic review and meta-analysis. HPB (Oxford). 2019;21(2):133-147. doi: 10.1016/j.hpb.2018.11.004 [DOI] [PubMed] [Google Scholar]
- 26.Pinheiro RS, Waisberg DR, Nacif LS, et al. Living donor liver transplantation for hepatocellular cancer: an (almost) exclusive Eastern procedure? Transl Gastroenterol Hepatol. 2017;2:68. doi: 10.21037/tgh.2017.08.02 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Iesari S, Inostroza Núñez ME, Rico Juri JM, et al. Adult-to-adult living-donor liver transplantation: the experience of the Université Catholique de Louvain. Hepatobiliary Pancreat Dis Int. 2019;18(2):132-142. doi: 10.1016/j.hbpd.2019.02.007 [DOI] [PubMed] [Google Scholar]
- 28.Rössler F, Sapisochin G, Song G, et al. Defining benchmarks for major liver surgery: a multicenter analysis of 5202 living liver donors. Ann Surg. 2016;264(3):492-500. doi: 10.1097/SLA.0000000000001849 [DOI] [PubMed] [Google Scholar]
- 29.Cheah YL, Simpson MA, Pomposelli JJ, Pomfret EA. Incidence of death and potentially life-threatening near-miss events in living donor hepatic lobectomy: a world-wide survey. Liver Transpl. 2013;19(5):499-506. doi: 10.1002/lt.23575 [DOI] [PubMed] [Google Scholar]
- 30.Gorgen A, Goldaracena N, Zhang W, et al. Surgical complications after right hepatectomy for live liver donation: largest single-center Western world experience. Semin Liver Dis. 2018;38(2):134-144. doi: 10.1055/s-0038-1636932 [DOI] [PubMed] [Google Scholar]
- 31.Choo SP, Tan WL, Goh BKP, Tai WM, Zhu AX. Comparison of hepatocellular carcinoma in Eastern versus Western populations. Cancer. 2016;122(22):3430-3446. doi: 10.1002/cncr.30237 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Missing Data Observed in the Databases Investigated
eTable 2. Upper Limit of Transplantability in the Different Centers
eFigure. Deceased- and Living-Donor Liver Transplantation Activity in the 12 Centers Composing the International Cohort