Hepatocellular carcinoma (HCC) is the most common primary malignant tumor of the liver and is the ninth leading cause of cancer deaths in the United States.1 Within the United States, certain ethnic groups have significantly increased incidence of HCC. In our practice area, the incidence of HCC for South Texas Latinos is 10.6 per 100 000.2 Additionally, the incidence of HCC will continue to escalate and is predicted to be the number 1 cause of cancer death in the United States by 2030.3 From an oncological and functional aspect, liver transplantation may be the optimal treatment for HCC in patients with end-stage liver disease.1
In 1996, Mazzaferro et al4 reintroduced the concept of transplantation for HCC by proposing the Milan criteria (single tumor <5 cm in size or <3 tumors each <3 cm in size). Their study showed a 4-year patient and tumor-free survival of 75% and 83%, respectively.4 In 2001, Yao et al5,6 expanded the Milan criteria introducing the University of California San Francisco (UCSF) criteria (solid tumors <6.5 cm in size or up to 3 tumors with <4.5 cm as a largest diameter and a total tumor burden of 8 cm) with similar survival results and in 2007 these results were validated. This led to a call for expansion of HCC criteria by Duffy et al7 in 2007. Since then, others, such as the Kyoto criteria developed in 2009 (tumor size up to 5 cm, up to 10 tumor nodules and des-γ-carboxy prothrombin values <400 mAU/mL), have continued to try to expand the eligibility criteria.8 More recently, the Milan group proposed the Metroticket Project creating an up to 7 criteria in tumors without vascular invasion with excellent 5-year survival (71.2%).9 These survival outcomes have been validated in multiple series.10,11 However, extending criteria beyond the Milan or UCSF risks may increase the risk of tumor recurrence and enhance concerns of reduced survival. Yet, recent large series have shown similar survival for patients selected by a variety of criteria as those compared to Milan.12
The dilemma remains that there is a limited supply of donated organs despite the benefit of transplantation. Currently, only 4%–6% of patients diagnosed with HCC are offered transplantation in Western countries.13 Chapman et al3 recently published their experience with downstaging large tumors into Milan criteria for transplantation. Although they had only 30% of the patients eligible for downstaging, they demonstrated survival outcomes similar to those who had initial tumor burden within Milan.3
At our institution, patients often present with large tumor burdens. To treat these patients, we created a comprehensive treatment plan for patients with HCC awaiting transplantation who are outside of UCSF criteria when indicated by our protocol. We created a protocol that employed tumor stability over extended time periods to define potential candidates. Our treatment regimen includes aggressive locoregional therapy (LRT), use of sorafenib in any patient demonstrating progression, and the posttransplant utilization of the mTOR inhibitor everolimus. We performed a retrospective review to determine the survival of patients outside UCSF criteria, inside UCSF but outside Milan, and inside Milan criteria.
MATERIALS AND METHODS
A retrospective review of 220 patients who underwent liver transplantation at The Sherrie and Alan Conover Center for Liver Disease and Transplantation at Houston Methodist Hospital, Houston, TX, between April 2008 and June 2017. All subjects were recruited and the procedures followed were in accordance with the ethical standards of the responsible Helsinki Declaration and the Declaration of Istanbul. Inclusion criteria were age >18, pretransplant diagnosis of HCC, no evidence of extrahepatic disease, and histologically proven HCC in the explant liver. Four recipients who had mixed tumors or multiple organ involvement on explant pathology were excluded. One of these patients was outside of Milan criteria. Five patients with incidental tumors were included in the pathologic analysis but excluded from the radiographic analysis. Radiographic tumor size and numbers were determined by multidisciplinary review of individual scans. The patients were separated into the following 3 groups: those having tumors meeting Milan criteria, 150 (70.8%); those outside Milan but inside UCSF criteria, 27 (12.7%); and those outside UCSF criteria, 35 (16.5%). Eligible patient records were reviewed, and data were collected for demographic characteristics, comorbid conditions, pre liver transplantation α-fetoprotein (AFP) levels, and type of tumor therapy. The Liver Imaging Reporting and Data System criteria were used to identify the pretransplant tumor presence and size.14
The patients were selected for orthotopic liver transplant (OLT) based on radiographic imaging and multidisciplinary imaging review. The patients would be considered for transplant within Milan by standard criteria. For patients outside of Milan, we developed a protocol so those who had control of their tumor would be considered for transplantation. Tumor control was defined as evidence of stable or improving size after LRT based on radiographic review at 3-month intervals. Patients with radiographic evidence of tumor thrombus or extrahepatic disease were excluded. Patients who had tumor expansion after LRT would undergo further therapy and could be reconsidered for transplantation if they exhibited tumor control as manifested by repeated imaging showing stable or improved tumor burden as well as temporal stability after repeat LRT. Before 2014, tumor stability was defined by the multidisciplinary committee on and individual basis. After 2014, a strict 9-month policy from the last LRT was used to define tumor stability. AFP was monitored; however, no absolute value of exclusion was used.
Based on the College of American Pathologists protocol, tumor number, maximum focal diameter, presence of vascular invasion, tumor differentiation, and the extent of HCC at the time of explant were examined.15,16 Using explanted tumor sizes, we separated the patients into 3 groups: those having tumors meeting Milan criteria, 138 (62.7%); those outside Milan but inside UCSF criteria, 23 (10.5%); and those outside UCSF criteria, 59 (26.8%). Patients were analyzed based on explant pathology rather than pretransplant radiographic size to help prevent underestimation of actual tumor burden seen with radiographic staging and allow radiographic misidentification to bias our cohort. Analysis of patients comparing radiographic and pathologic stratification was done to ensure pretransplant radiographic assessments were adequate in selecting patients with higher tumor burden.
Patients were followed in the standard fashion for our institution using a multidisciplinary transplant team. Primary outcomes were patient survival at 1, 3, and 5 years after OLT. Disease-free survival (DFS) was defined as posttransplant tumor recurrence identified by either histological and/or radiological evidence of HCC recurrence. Recipients were placed on surveillance imaging per our programmatic protocol with imaging every 6 months for 3 years and then annually. Recipients alive or lost to follow-up were censored at the date of last follow-up.
Demographic and clinical data were reported as frequencies and proportions for categorical variables and as median and interquartile range (IQR) for continuous variables. Difference across groups was compared using the χ2 or Fisher exact tests for categorical variables and Kruskal-Wallis test for continuous variables. Overall, patient survival and HCC-free survival were presented by the Kaplan-Meier curves. Differences across groups were compared using the log-rank test. Univariate survival analysis and multivariate Cox proportional hazards modeling were performed to determine the characteristics associated with the risk of HCC recurrence within 5 years from transplant. Variables having P < 0.2 in the univariate analysis or variables of clinical significance were investigated further in multivariate Cox proportional hazards modeling. Variable selection for the Cox proportional hazards models was conducted using the Bayesian model averaging method. The best model was selected based on the smallest Bayesian information criterion. All the analyses were performed on Stata version 14.2 (StataCorp LLC, College Station, TX). A P < 0.05 was considered statistically significant.
Local Regional Therapy
LRT consisted of treatment modalities that were not surgical resection. Surgical interventions were categorized separately and did not determine total tumor burden. Locoregional therapies included: transarterial chemoembolization (TACE), yttrium-90 therapy (both segmentectomy and lobar therapies), and radiofrequency ablation. The modality chosen for individual patients was made after review at multidisciplinary conference. Patient’s tumor location, size, and imaging appearance determined the modality chosen for treatment. Patients were reviewed every 3 months per protocol to assess for response to LRT.
Sorafenib, a multikinase inhibitor Food and Drug Administration approved in advanced HCC, was initiated in our patients where there was evidence of disease progression after LRT.17 Disease progression was defined as radiographic extension of a tumor or residual tumor after local regional therapy. Determination of disease progression was made via multidisciplinary review of computed tomographic or MRI scans. Sorafenib dosing was started at 200 mg twice a day and titrated in 2 weeks to 400 mg twice per day depending on tolerability. Discontinuation of medication was done at the time of transplantation or due to intolerance. Due to the heterogeneity of the individual data for dosing and duration of therapy, patients given any sorafenib were included in analysis.
As shown in Table 1, the median age of the study population was 61 years. The patients were largely male (72.3%). Hepatitis C virus was present as underlying liver disease in 69.5%. The median biological Model for End-stage Liver Disease (MELD) was 13.0 and MELD by exception was 29. The median waitlist time was 344 days (IQR, 196–522). The vast majority of recipients (93.2%) received LRT before transplant. TACE was the predominant form of LRT being used in 82.7% of the patients. Radiofrequency ablation and yttrium-90 were also used as LRT. LRT was more commonly used in the patients with larger tumor burden. Milan patients having on average 1.4 (±0.8) total LRT while UCSF using 2.1 (±1.4), and outside UCSF requiring 2.5 (±1.5) (P < 0.001). The larger tumor groups had received more LRT before listing and after listing as well. After listing the Milan group required 0.8 (±0.8) LRT with UCSF requiring 1.2 (±0.9) and outside UCSF 1.6 (±1.4) (P < 0.01).
Sorafenib was utilized in 57 of the 220 patients (25.65%). Median waiting time was significantly different between the groups with Milan group waiting 376 days (IQR, 210–532) days, UCSF patients waiting 199 days (IQR, 110–370), and outside UCSF patients waiting 307 days (IQR, 182–538) (P = 0.01).
Patients within Milan, UCSF, and beyond UCSF largely exhibited similar demographic and clinical characteristics (Table 2). Differences did exist with patients inside Milan having a lower biologic MELD score at OLT (P = 0.049). The MELD score at the time of offer was similar between the groups. The AFP at the time of listing was lower in the Milan group (8.4) than the UCSF group (22.4) or the outside UCSF group (16.7) (P = 0.004). This same pattern was noted with the last AFP before transplant with groups showing values of 6.0 versus 17.3 versus 12.1, respectively (P < 0.001). Only 11.9% (7/59) of patients outside UCSF underwent resection before transplantation. These patients had resection with subsequent evidence of either recurrence of primary tumor or de novo HCC in the remnant liver. Significantly more patients 30/59 (50.8%) outside UCSF received sorafenib before OLT compared to 13.8% inside Milan and 34.8% inside UCSF criteria (P < 0.001).
Overall, patient survival was not different in the 3 groups. The 1-year survival for the Milan cohort was 92%, UCSF 100%, and beyond UCSF 97%. Three-year survival was also not significantly different with the groups showing 87%, 88%, and 87%, respectively. Even 5-year survival was closely mirrored in all groups with 81%, 88%, and 80% survivals (Figure 1). DFS was noted to be similar among groups with 1-year survival for Milan 100%, inside UCSF 95.5%, and outside UCSF at 91.1%. DFS at 5 years revealed 92% survival for Milan, 88.6% inside UCSF, and 85.4% outside UCSF (P = 0.53; Figure 2).
Before transplant, there were 150 patients meeting Milan criteria, 27 in UCSF criteria, and 35 outside UCSF criteria. These patients were regrouped based on explantation tumor burden after transplantation. We performed a similar analysis on radiographic categorization of these patients and found that they exhibited largely the same characteristics as when grouped by pathology. Comparing the radiographic to the pathologic tumor burden, there was a 72% agreement (P < 0.001) between radiographic and pathologic classification with a Cohen κ coefficient of 0.44 (P < 0.001). The 1-year survival for the radiographic Milan cohort was 94.4%, UCSF 96.3%, and beyond UCSF 100%. Three-year survival was also not significantly different with the groups showing 83.4%, 81.6%, and 90.6%, respectively. Even 5-year survival was closely mirrored in all groups with 77.0%, 81.6%, and 70.5% survivals (P = 0.95). DFS was different when classified by radiographic criteria in our series. With 5-year DFS of 93.4% in Milan patients, 77.0% in UCSF, and 87.3 % in patients outside UCSF (P = 0.009). However, when comparing each group to one another, there was no difference in patient survival or DFS.
The significant pathological differences between the 3 groups including number and size (P ≤ 0.001) are shown in Table 3. Tumor differentiation was statistically significant (P < 0.01) between the 3 cohorts with great numbers of poorly differentiated tumors noted in the outside UCSF group. Microvascular invasion in the outside UCSF group was significantly different as compared to the other groups. Microvascular invasion was noted in 4.3% within the Milan cohort, 8.7% within UCSF, and 18.6% outside UCSF criteria (P = 0.01). In over 26.1% of the Milan patients, no viable tumor was noted on explant. Viable tumor was noted more often in the patients outside Milan criteria (P < 0.001). Tumor necrosis percentage was determined by review of the pathology by the pathologist MD. Increased necrosis was noted in the Milan and UCSF groups as compared to the outside UCSF group (80 versus 80 versus 65; P = 0.03). Recurrence of tumor was noted to be higher in the larger tumor groups with 3.6% recurrence noted in Milan group, 8.7% in the UCSF group, and 15.3% in the outside UCSF group (P = 0.02). Time to recurrence and recurrence location was not different among the groups.
Predictors of Recurrence
Univariate analysis demonstrated the variables predicting HCC recurrence were tumor size (P = 0.01), microvascular invasion (P < 0.001), poor tumor differentiation based on pathologic criteria (P = 0.003), recipients body mass index (BMI) at OLT (P = 0.04), recipients outside UCSF criteria (P = 0.02), and AFP immediately pre-OLT >100 (P = 0.04) (Table 4). Multivariate analysis demonstrated tumor microvascular invasion (P < 0.001) and BMI at OLT (P = 0.04) were the only independent risk factors for HCC recurrence, while poor tumor differentiation approached significance (HR=4.62; P=0.08). HCC recurrence rates were significantly different when comparing patients within Milan criteria (3.6%) and those outside of UCSF (15.3%) (P = 0.02). When using radiographic stratification of tumors for multivariate analysis, similar factors were elicited. However, poor tumor differentiation was the only additional significant factor when compared with pathologic analysis (P = 0.01). AFP was not used for selection of patients for transplant. In our multivariate analysis, AFP was not a predictor of recurrence in our cohort.
During the initiation of this program, an interim analysis was done to try and determine if changes to patient selection were needed. In 2014, we reviewed our data and noted 21 patients with tumors too large for MELD exception who were removed from the transplant list due to disease progression. Eighteen of 21 (86%) dropped out because of progressive disease by 9 months (median time of drop out 197 days). We have adopted a policy of radiographic tumor stability of 9 months since that analysis. Since 2014 there have been 68 patients delisted who had a diagnosis of HCC. The average time to delisting is 296.7 days (median 260). For patients who are listed with out of standard criteria tumor burden since 2014, the average time to removal from the transplant list was 215.5 days. The median of 52 days reflects that patients were often removed quite earlier from the list but one patient was removed after 1161 days. Patients with a waiting time of >9 months demonstrating radiologic stability did not display recurrence rates that were different from those within Milan or UCSF criteria. In contrast, patients with a waiting time <9 months had significantly higher rates of recurrence (P = 0.046) (Table 5). Overall in the 16 patients with recurrence, 10 (62.5%) had a median waiting time on the list of 121 days (range, 88–249), whereas 6 (37.5%) had a median waiting time of 510 days (range, 351–1131) (P = 0.001). Since adopting a 9-month waiting period in 2014, recurrence rates have not changed significantly (P = 0.07). When the patients were grouped by pretransplant radiographic tumor burden, a waiting time of <9 months was an also predictor of recurrence (P = 0.03). The time on the waiting list has increased significantly for patients transplanted after 2014 (P < 0.001) (Table 6). This is also true for patients stratified by radiographic tumor burden (P < 0.001).
Liver transplantation is the best current curative option for patients with HCC. Larger tumors are known to reappear after transplant at a higher rate.4 However, not all large tumors will recur. We presented a robust single-center experience of transplantation for large tumors outside of Milan and UCSF criteria. In this 5-year retrospective study, overall survival rates for patients with tumors inside Milan, outside Milan but inside UCSF and outside UCSF criteria were equivalent. The patients were stratified by pathologic tumor burden due to concerns for tumor recurrence in large tumors as well as the underestimation of tumor burden by radiographic measure.4 We intended to show that even when using a larger number of actual tumor burden, patients carefully selected using time as a variable can achieve equivalent outcomes with those within Milan criteria tumors. There was favorable concordance when using pathologic or radiographic characterization of tumor.
We experienced a 7.2% (n = 16) overall recurrence rate. Nine of the 16 recurrences occurred in the outside UCSF patients who experienced an overall recurrence rate of 15.3% within that subgroup. It was significantly higher than in the patients with tumors inside Milan or inside UCSF criteria (P = 0.02). Yet, this rate of recurrence is lower than in other series.7 It is possible this lower rate of recurrence is the temporal criteria we applied for patient selection. In 2014, we identified a time threshold of 9 months in which 86% (18/21) of patients dropped out due to progression before transplantation. Overall, 7 patients (70%) outside Milan criteria had recurrences and were transplanted before this 9-month threshold. Previously, Samoylova et al15 identified time on the list (<120 days) as a predictor of HCC recurrence. Yao et al6 had a median time to transplant (LRT and OLT) of 8.2 months with 30% dropout rate due to tumor progression. Explant histology revealed favorable qualities (better tumor differentiation and less microvascular invasion) suggesting that perhaps waiting time selected eliminated patients with unfavorable tumor biology.6 In our 16 patients with recurrence, 10 (62.5%) had a median waiting time on the list of 121 days (range, 88–249), whereas 6 (37.5%) had a median waiting time of 510 days (range, 351–1131) (P = 0.001). Roberts et al18 when describing the ablate and wait strategy for transplantation stated that, “time may be the surest method for selecting patients with HCC for transplantation destined to have good outcomes.”18 Recent adoption of changes in HCC prioritization to incorporate a waiting time of at least 6 months is further evidence that time continues to likely be the best current marker of tumor biology. Mehta et al19 recently published the experience of 911 patients considered for MELD exception from 2002 to 2012. They concluded that a sweet spot existed from 6 to 18 months after diagnosis of HCC where there was decreased risk of recurrence.19 In contrast, a large single-center experience from Palmer et al20 showed no change in survival or recurrence in patients transplanted before or after 180 days.20 We had adopted a stability of 9 months after LRT that may help to define tumor biology. Thus, time to transplantation may be one of the best criteria to select patients for OLT, regardless of tumor size.
Bridging therapies and downstaging have been shown to decrease patient drop out and increase survival posttransplantation for HCC.21,22 Bridging therapies include TACE, radiofrequency ablation (RFA), and resection. Downstaging can potentially achieve a reduction in tumor progression, volume, number, and even have a positive impact on post-LT recurrence rates.3 Patients meeting UNOS exception criteria (T2) who may have a waiting time >6 months should be considered for LRT.23 The effect of locoregional therapies impact on preventing waitlist drop out has yielded mixed results.24-28 A recent meta-analysis showed that for T2 lesions, bridging therapy showed a “non-significant reduction in the risk of waitlist dropout due to progression RR 0.32 (95% CI 0.06-1.85, I2 = 0%).”29
In the current study, HCC recurrence occurred more frequently in the first 2 years with median time to recurrence of 14-month posttransplantation, which is consistent with previous studies.1,30-32 The majority of patients (93%) received aggressive preoperative LRT and 13.8% of patients received sorafenib. Eleven (5.0%) out of the 220 recipients had tumor resection. As expected, recipients with explant tumors outside UCSF criteria (high risk) had the higher HCC recurrence incidences (15.3%). This was significantly different compared to patients with smaller tumors. In multivariate analysis, microvascular invasion was the strongest predictor of recurrence (P < 0.001). We were somewhat surprised to see the lower BMI was a significant predictor of tumor recurrence. This has not been seen in other series. Its significance is unclear but there is possible some protective effect for increase adiposity in tumor recurrence. However, the sample size here is too small to draw any definitive conclusions.
Noninvasive predictive features for recurrence after transplantation have been tumor size and number, radiologic evidence of microvascular invasion, and metastasis. Other studies have also suggested that the tumor markers AFP, AFP-L3, and des-γ-carboxy prothrombin may have predictive value.33-36 Some studies have suggested cut off values from 300 ng/mL to 1000 ng/mL as criteria for delisting patients, however, the most significant determinant appears to be a steadily rising AFP level >15 ng/mL per month.37 Tumor biology can likely better be predicted by biopsy; however, the morbidity risk outweighs its benefit. These risks can include tumor seeding, false negative results, vascular invasion which may occur after biopsy, and ultimately the aggressiveness of the tumor better assessed during explant examination.38-40 Recent data have identified possible biomarkers that intend to predict the behavior of HCC.17,41-43 The possibility of successfully identifying patients who are ideal candidates for transplantation despite tumor size is the ideal situation. To date, none of these markers have consistently helped to determine who will respond to LRT, chemotherapies, or identified those at low risk of recurrence.
In the current study, univariate analysis found an AFP >100 directly before transplant to be a significant predictor, though not seen on multivariate. We did note that 8 patients had an AFP >1000, yet only one recurrence was noted in this group. In our multivariate analysis, AFP was not a predictor of recurrence in our cohort. Vascular invasion appeared to be the strongest independent predictor affecting HCC recurrence in multivariate model. The other significant aspects (presence of vascular invasion, poor tumor differentiation, and explant tumor outside UCSF criteria) all were obtained with explant pathology. Unfortunately, these are very difficult to assess pretransplantation. Thus, our study further highlights the need for a reliable peripheral marker for HCC behavior.
Our study has some limitations, namely in the retrospective nature, as this is a single-center study and this study involved a relatively small number of patients. Furthermore, LRT and sorafenib used before transplantation were based on clinical progression determined by our multidisciplinary group rather than strict criteria. Ultimate tumor size was based on explant pathology rather than pretransplant imaging. Prior studies found no significant differences in patient and recurrence-free survivals between pre- and post-transplant classifications.7,44 Pretransplant staging failed to predict the actual number of HCC tumors at pathology in about 1/3 of patients and only largest nodule diameter predicted recurrence after LT.15,50 A previous study demonstrated that ≈25% of patients within the Milan and 28% within the UCSF criteria were underestimated by the pretransplant imaging.45 We performed analysis based on radiographic stratification in the same patients and there was good concordance of data. We felt is necessary to ensure that even with radiographic tumor selection that these results would hold up when the pathology was examined. We did note that a waiting time <9 months was a more common occurrence in patients who recurred when using radiographic criteria. Despite these limitations, this is a large series that shows successful transplantation of large tumors outside of UCSF criteria. The series further highlights the need for a combination approach using LRT and temporal stability to select biologically favorable tumor biology.
Liver transplantation is a curative therapeutic modality for selected patients with HCC and end-stage liver diseases. We reported a large series of successful transplantation in patients beyond UCSF criteria. In our series, patient survival and recurrence-free survival was equivalent to patients within Milan criteria. Our experience suggests that a waitlist time >9 months with aggressive LRT may help to define candidates outside of traditional size criteria who would have favorable outcomes with transplant. Until a reliable biomarker is available, temporal assessment of tumor biology in patients outside UCSF criteria is likely the most reliable predictor.
1. Kornberg A, Küpper B, Tannapfel A, et al. Long-term survival after recurrent hepatocellular carcinoma in liver transplant patients: clinical patterns and outcome variables. Eur J Surg Oncol. 2010; 36:275–280
2. El-Serag HB, Kanwal F. Epidemiology of hepatocellular carcinoma in the United States: where are we? Where do we go? Hepatology. 2014; 60:1767–1775
3. Chapman WC, Garcia-Aroz S, Vachharajani N, et al. Liver transplantation for advanced hepatocellular carcinoma after downstaging without up-front stage restrictions. J Am Coll Surg. 2017; 224:610–621
4. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med. 1996; 334:693–699
5. Yao FY, Xiao L, Bass NM, et al. Liver transplantation for hepatocellular carcinoma: validation of the UCSF-expanded criteria based on preoperative imaging. Am J Transplant. 2007; 7:2587–2596
6. Yao FY, Kerlan RK Jr, Hirose R, et al. Excellent outcome following down-staging of hepatocellular carcinoma prior to liver transplantation: an intention-to-treat analysis. Hepatology. 2008; 48:819–827
7. Duffy JP, Vardanian A, Benjamin E, et al. Liver transplantation criteria for hepatocellular carcinoma should be expanded: a 22-year experience with 467 patients at UCLA. Ann Surg. 2007; 246:502–9. discussion 509
8. Fujiki M, Takada Y, Ogura Y, et al. Significance of Des-gamma-carboxy prothrombin in selection criteria for living donor liver transplantation for hepatocellular carcinoma. Am J Transplant. 2009; 9:2362–2371
9. Mazzaferro V, Llovet JM, Miceli R, et al; Metroticket Investigator Study Group. Predicting survival after liver transplantation in patients with hepatocellular carcinoma beyond the Milan criteria: a retrospective, exploratory analysis. Lancet Oncol. 2009; 10:35–43
10. Gugenheim J, Bredt LC, Iannelli A, et al. Recurrence after liver transplantation for hepatocellular carcinoma according to up-to-seven criteria. Hepatogastroenterology. 2013; 60:799–806
11. Raj A, McCall J, Gane E. Validation of the “Metroticket” predictor in a cohort of patients transplanted for predominantly HBV-related hepatocellular carcinoma. J Hepatol. 2011; 55:1063–1068
12. Xu X, Lu D, Ling Q, et al. Liver transplantation for hepatocellular carcinoma beyond the Milan criteria. Gut. 2016; 65:1035–1041
13. Mazzaferro V, Battiston C, Sposito C. Pro (with caution): extended oncologic indications in liver transplantation. Liver Transpl. 2018; 24:98–103
14. Mitchell DG, Bruix J, Sherman M, et al. LI-RADS (liver imaging reporting and data system): summary, discussion, and consensus of the LI-RADS management working group and future directions. Hepatology. 2015; 61:1056–1065
15. Samoylova ML, Dodge JL, Yao FY, et al. Time to transplantation as a predictor of hepatocellular carcinoma recurrence after liver transplantation. Liver Transpl. 2014; 20:937–944
16. Shah SA, Tan JC, McGilvray ID, et al. Accuracy of staging as a predictor for recurrence after liver transplantation for hepatocellular carcinoma. Transplantation. 2006; 81:1633–1639
17. Shao YY, Hsu CH, Cheng AL. Predictive biomarkers of sorafenib efficacy in advanced hepatocellular carcinoma: are we getting there? World J Gastroenterol. 2015; 21:10336–10347
18. Roberts JP, Venook A, Kerlan R, et al. Hepatocellular carcinoma: ablate and wait versus rapid transplantation. Liver Transpl. 2010; 16:925–929
19. Mehta N, Heimbach J, Lee D, et al. Wait time of less than 6 and greater than 18 months predicts hepatocellular carcinoma recurrence after liver transplantation: proposing a wait time “sweet spot”. Transplantation. 2017; 101:2071–2078
20. Palmer WC, Lee D, Burns J, et al. Liver transplantation for hepatocellular carcinoma: impact of wait time at a single center. Ann Hepatol. 2017; 16:402–411
21. Ravaioli M, Grazi GL, Piscaglia F, et al. Liver transplantation for hepatocellular carcinoma: results of down-staging in patients initially outside the Milan selection criteria. Am J Transplant. 2008; 8:2547–2557
22. Cescon M, Ravaioli M, Grazi GL, et al. Prognostic factors for tumor recurrence after a 12-year, single-center experience of liver transplantations in patients with hepatocellular carcinoma. J Transplant. 2010; 2010
23. Graziadei IW, Sandmueller H, Waldenberger P, et al. Chemoembolization followed by liver transplantation for hepatocellular carcinoma impedes tumor progression while on the waiting list and leads to excellent outcome. Liver Transpl. 2003; 9:557–563
24. Hayashi PH, Ludkowski M, Forman LM, et al. Hepatic artery chemoembolization for hepatocellular carcinoma in patients listed for liver transplantation. Am J Transplant. 2004; 4:782–787
25. Maddala YK, Stadheim L, Andrews JC, et al. Drop-out rates of patients with hepatocellular cancer listed for liver transplantation: outcome with chemoembolization. Liver Transpl. 2004; 10:449–455
26. Fontana RJ, Hamidullah H, Nghiem H, et al. Percutaneous radiofrequency thermal ablation of hepatocellular carcinoma: a safe and effective bridge to liver transplantation. Liver Transpl. 2002; 8:1165–1174
27. Lu DS, Yu NC, Raman SS, et al. Percutaneous radiofrequency ablation of hepatocellular carcinoma as a bridge to liver transplantation. Hepatology. 2005; 41:1130–1137
28. Otto G, Herber S, Heise M, et al. Response to transarterial chemoembolization as a biological selection criterion for liver transplantation in hepatocellular carcinoma. Liver Transpl. 2006; 12:1260–1267
29. Kulik L, Heimbach JK, Zaiem F, et al. Therapies for patients with hepatocellular carcinoma awaiting liver transplantation: a systematic review and meta-analysis. Hepatology. 2018; 67:381–400
30. Yokoyama I, Carr B, Saitsu H, et al. Accelerated growth rates of recurrent hepatocellular carcinoma after liver transplantation. Cancer. 1991; 68:2095–2100
31. Frenette CT, Boktour M, Burroughs SG, et al. Pre-transplant utilization of sorafenib is not associated with increased complications after liver transplantation. Transpl Int. 2013; 26:734–739
32. Hoffmann K, Ganten T, Gotthardtp D, et al. Impact of neo-adjuvant sorafenib treatment on liver transplantation in HCC patients - a prospective, randomized, double-blind, phase III trial. BMC Cancer. 2015; 15:392
33. Ma WJ, Wang HY, Teng LS. Correlation analysis of preoperative serum alpha-fetoprotein (AFP) level and prognosis of hepatocellular carcinoma (HCC) after hepatectomy. World J Surg Oncol. 2013; 11:212
34. Herrero JI, Sangro B, Quiroga J, et al. Influence of tumor characteristics on the outcome of liver transplantation among patients with liver cirrhosis and hepatocellular carcinoma. Liver Transpl. 2001; 7:631–636
35. Parfitt JR, Marotta P, Alghamdi M, et al. Recurrent hepatocellular carcinoma after transplantation: use of a pathological score on explanted livers to predict recurrence. Liver Transpl. 2007; 13:543–551
36. Sapisochin G, Facciuto M, Rubbia-Brandt L, et al; iCCA International Consortium. Liver transplantation for “very early” intrahepatic cholangiocarcinoma: international retrospective study supporting a prospective assessment. Hepatology. 2016; 64:1178–1188
37. Zhao WC, Fan LF, Yang N, et al. Preoperative predictors of microvascular invasion in multinodular hepatocellular carcinoma. Eur J Surg Oncol. 2013; 39:858–864
38. Lopez KT, Kuwada SK, Wong LL. Consequences of needle tract seeding of hepatocellular cancer after liver transplant. Clin Transplant. 2013; 27:E400–E406
39. Marsh JW, Dvorchik I, Subotin M, et al. The prediction of risk of recurrence and time to recurrence of hepatocellular carcinoma after orthotopic liver transplantation: a pilot study. Hepatology. 1997; 26:444–450
40. Clavien PA, Lesurtel M, Bossuyt PM, et al. Recommendations for liver transplantation for hepatocellular carcinoma: an international consensus conference report. Lancet Oncol. 2012; 13:e11–e22
41. Ashida R, Okamura Y, Ohshima K, et al. CYP3A4 gene is a novel biomarker for predicting a poor prognosis in hepatocellular carcinoma. Cancer Genomics Proteomics. 2017; 14:445–453
42. Nörthen A, Asendorf T, Walson PD, et al. Diagnostic value of alpha-1-fetoprotein (AFP) as a biomarker for hepatocellular carcinoma recurrence after liver transplantation. Clin Biochem. 2018; 52:20–25
43. Shao YY, Lin H, Li YS, et al. High plasma interleukin-6 levels associated with poor prognosis of patients with advanced hepatocellular carcinoma. Jpn J Clin Oncol. 2017; 47:949–953
44. Grasso A, Stigliano R, Morisco F, et al. Liver transplantation and recurrent hepatocellular carcinoma: predictive value of nodule size in a retrospective and explant study. Transplantation. 2006; 81:1532–1541
45. Bonadio I, Colle I, Geerts A, et al. Liver transplantation for hepatocellular carcinoma comparing the Milan, UCSF, and Asan criteria: long-term follow-up of a Western single institutional experience. Clin Transplant. 2015; 29:425–433