Hepatocellular carcinoma (HCC) has become the fastest growing cause of mortality among all types of cancer in the United States.1 Most HCC (>70%) occurs with a background of cirrhosis. Liver resection in patients with cirrhosis is associated with high morbidity and mortality. Liver transplantation has clearly been established as the primary treatment option in patients with HCC and cirrhosis as it can provide adequate oncological resection as well as treatment for the underlying liver disease. Since the first success with living donor adult liver transplantation (LDALT) in 1998, LDALT has become a significant therapeutic modality in the treatment of HCC, particularly in centers in which access to deceased donor organs is limited. This is evident in a number of countries in Asia and the Middle East and even in some regions in the United States. Under the current United Network for Organ Sharing (UNOS) system, the median time to transplantation for all patients in the United States with stage 2 HCC who received exception points was 54 days in 2005.2 Our institution is located in Organ Procurement and Transplantation Network Region 1, in which the median time to transplantation for patients with HCC in 2006 was 168 days.2 Patients without HCC and Model for End-Stage Liver Disease (MELD) scores of 20 to 29 had a median time to transplantation of just over 275 days in our region.2 This difference in the wait time to transplantation has rendered LDALT a major therapeutic modality for transplant candidates who will otherwise have a higher wait-list mortality rate because of longer waiting times. Centers in Japan and Korea have reported large experiences with living donor liver transplantation for HCC, demonstrating 3-year survival rates of 69% and 73% in their patients.3, 4
In comparison with Asian centers, there is far less experience with LDALT in the United States. One multicenter study [the Adult-to-Adult Living Donor Liver Transplantation (A2ALL) cohort study] reported a 3-year survival rate of 67% and a tumor recurrence rate of 29%.5 The high rate of HCC recurrence dampened the authors' enthusiasm for LDALT as a treatment for HCC. Because of the dependence of our patient population on LDALT due to the organ scarcity in our region, these findings prompted us to assess the outcomes of our patients who underwent LDALT for HCC. The main aim of this study was to assess tumor recurrence and overall survival in our patient population. In addition, we examined the effects of various tumor and patient characteristics on recurrence and survival.
A2ALL, Adult-to-Adult Living Donor Liver Transplantation; AFP, alpha fetoprotein; DDLT, deceased donor liver transplantation; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; LDALT, living donor adult liver transplantation; MELD, Model for End-Stage Liver Disease; N/A, not applicable; NASH, nonalcoholic steatohepatitis; UCSF, University of California San Francisco; UNOS, United Network for Organ Sharing.
PATIENTS AND METHODS
From October 1999 to August 2007, 139 LDALT procedures using right lobe grafts were performed at the Lahey Clinic Medical Center. Twenty-eight patients (20.1%) had HCC identified in the explanted livers. Of these, 22 patients had pretransplant imaging studies consistent with a diagnosis of HCC. The overall median follow-up after LDALT was 3.4 years (range = 0.25-8.7 years). The median follow-up of LDALT patients alive at the end of the study was 4.1 years (range = 1-8.7). The primary etiologies of cirrhosis in 27 patients were hepatitis C virus (n = 11), alcohol (n = 4), cryptogenic (n = 4), hepatitis C virus and alcohol (n = 4), hepatitis B virus (n = 2), nonalcoholic steatohepatitis (n = 1), and hemochromatosis (n = 1). One patient did not have cirrhosis but had undergone a partial right hepatectomy for HCC and a subsequent partial left hepatic resection for recurrent HCC. The male to female ratio was 21:7, and the median age at transplantation was 56 years (range = 47-67). In patients with a pretransplant diagnosis of HCC, the tumor size was evaluated with computed tomography scans or magnetic resonance imaging. Patients who were accepted as candidates for LDALT had tumors that were within the Milan criteria. The presence of extrahepatic disease was an absolute contraindication to transplantation. Alpha-fetoprotein levels were available in 26 patients prior to transplantation. Recipient and donor evaluation and selection as well as our operative techniques for LDALT have been previously described.6 Patient demographics are summarized in Table 1.
Table 1. Demographics of Living Donor Adult Liver Transplantation Recipients
Abbreviations: AFP, alpha fetoprotein; HBV, hepatitis B virus; HCV, hepatitis C virus; MELD, Model for End-Stage Liver Disease; N/A, not applicable; NASH, nonalcoholic steatohepatitis.
Mean age at transplant (range)
Etiology of cirrhosis
HCV plus alcohol
MELD score (without tumor points)
All recipients until August 2008 received induction immunosuppression with daclizumab (1 mg/kg on postoperative days 0 and 4). All recipients received steroids in the anhepatic phase. The standard maintenance immunosuppression regimen was based on tacrolimus (target levels = 8-12 mg/L), mycophenolate mofetil, and a prednisone taper. Over the past year, we have transitioned to a 5-day steroid regimen, at the end of which patients are completely off steroids.
All pathology slides of the explanted livers in the LDALT group were re-examined by the same experienced liver pathologist. Vascular invasion refers to microscopic infiltrates as no macroscopic infiltrations were noted on pretransplant imaging or on the explanted livers.
Data Collection and Statistical Analysis
The study was approved by the hospital's institutional review board. Patient data were obtained from a prospectively maintained computerized database and medical records. Data were analyzed with the SPSS statistical software package. Survival data were analyzed with the Kaplan-Meier method and the log-rank statistic. Categorical comparison was performed with the chi-square test. Statistical analysis was set at P < 0.05 for all tests.
Of 28 patients, 22 (78.6%) had pretransplant imaging consistent with a diagnosis of HCC. The mean explant tumor size was 3.3 ± 1.2 cm, and the mean number of tumors was 1.5 ± 0.8. The mean sum of all the tumors on explant pathology was 4.1 ± 2 cm. The mean size of the largest tumor was 3.4 ± 1.0 cm on pretransplant imaging studies. There was a statistically significant correlation (P < 0.005) between the size of the largest tumor on pretransplant imaging and the size of the tumor identified in the explanted liver. On the basis of explant pathology, 21 patients (75%) had tumors within the Milan criteria, 5 patients (17.9%) were outside the Milan criteria but were within the University of California San Francisco (UCSF) criteria, and 2 patients (7.1%) were beyond the UCSF criteria. Tumor differentiation was categorized as well in 5, moderate in 19, well to moderate in 1, and poor in 3 patients. Vascular invasion was present in 13 patients (46.4%). Histopathological data are summarized in Table 2.
Table 2. Tumor Characteristics in Explanted Livers of Living Donor Adult Liver Transplantation Recipients
Time of tumor diagnosis
Median tumor size (cm)
3.5 (range = 0.8-5)
Median tumor number
1 (range = 1-4)
Largest tumor size on explant
Well to moderately differentiated
The patient without a diagnosis of cirrhosis had undergone 2 previous partial liver resections for HCC. Radiofrequency ablation prior to transplantation was performed in 4 patients (14.3%), with 1 patient developing tumor recurrence at 2.3 years post-transplantation. The median follow-up in this pretreated group was 3.8 years (range = 2-7.3). There were no deaths in this pretreated group. The histological examination of tumors that had undergone RFA demonstrated 25% to 80% tumor necrosis. In these cases, we used the pretransplant size on imaging studies for our analysis because it was difficult to assess the exact volume of the viable tumor in the explanted liver. The median largest tumor size on preoperative imaging in this group was 3.5 cm (range = 1.3-4.6).
Approximately one-third of the liver donors developed a complication (8/28, 29%) as a result of right lobe donation. None of the patients had more than 1 complication. The complications included wound infection (2), biloma (2), urinary tract infection (2), intra-abdominal collection (1), and pulmonary edema (1). The complications were Clavien 1, 2, or 3.7 None of the complications were life-threatening. Two patients with a biloma required percutaneous drainage by interventional radiology. One patient with an intra-abdominal collection underwent percutaneous drainage for a presumed bile leak, but the drainage fluid was found to be a sterile, nonbilious collection.
Patient and Graft Survival
The overall 1- and 5-year patient survival rates following LDALT were 96% and 81%, respectively (Fig. 1). Graft survival was equivalent to patient survival because none of the patients developed graft failure. None of the LDALT recipients underwent retransplantation. In all, there were 4 deaths (14.3%) in the LDALT group during the study period. All the deaths were a result of recurrent HCC. The tumor grade had a statistically significant effect on survival (P < 0.05; Fig. 2). The overall 3-year patient survival rates for patients with well and moderately differentiated tumors were 100% and 91.7%, respectively. None of the patients with poorly differentiated tumors lived beyond 2.8 years. Patients with tumors within the Milan criteria had a 5-year survival rate of 87.1% versus 80% (P = 0.43; Fig. 3) for those patients exceeding the Milan criteria but within the UCSF criteria. Of the 2 patients with tumors beyond the UCSF criteria, 1 died of recurrent disease at 3.1 years, and the other is alive at 8.7 years, free of tumor recurrence. The tumor size, number of tumors, sum of tumors, alpha fetoprotein (AFP) level, age, and MELD score were not significant factors for patient survival on univariate analysis.
Eight patients (28.6%) developed tumor recurrence during the follow-up period. Areas of recurrence were the liver (2), lung (2), lymph nodes (2), bone (2), pancreas (1), and diaphragm (1). Some patients had more than 1 site of recurrence. In the majority of cases, the diagnosis of recurrence within the liver was based on typical findings on imaging studies. Patients with only extrahepatic disease underwent biopsy to confirm the diagnosis. The mean time to recurrence was 2.4 ± 1.6 years, and the mean length of survival from the time of diagnosis of recurrent HCC to the end of the follow-up period for those patients alive with recurrence was 2.0 ± 1.4 years. These patients had a mean time from the diagnosis of tumor recurrence to death of 0.7 ± 0.5 years, and 75% of them had poorly differentiated HCC. Tumor differentiation was a significant determinant of recurrence (P = 0.006), with 100% recurrence in patients with poorly differentiated HCC. None of the patients with well-differentiated tumors developed recurrence, and 26.7% of those with moderately differentiated tumors developed recurrence. Of the 8 patients with recurrence, 5 patients had microvascular invasion on histopathology, yet the presence of vascular invasion was not a statistically significant predictor of recurrence (P = 0.28).
The mean tumor size was larger (3.5 ± 0.7 versus 3.1 ± 1.3) in those patients with recurrence; however, the difference was not statistically significant (P = 0.44). There were no statistically significant differences in the AFP levels, number of tumors, or tumor size on imaging studies done prior to transplantation or in the total number of tumors, sum of the tumor size in the explanted liver, or MELD score at the time of transplantation in patients with recurrence in comparison with those without recurrence. Patients with recurrent HCC were significantly younger than those without tumor recurrence (mean age of 51.4 ± 3.7 versus 57.9 ± 5.7 years, P = 0.006).
HCC recurrence occurred in nearly a quarter of the patients (5/21, 23%) with tumors within the Milan criteria versus 42.8% (3/7) in those beyond the Milan criteria (P > 0.05). One of the 2 patients with tumors beyond the UCSF criteria developed recurrent HCC.
Overall, 4 patients are alive with recurrence; of these, 3 patients had moderately differentiated tumors, and 1 had a well to moderately differentiated tumor. The median time to recurrence in this group was 3.6 years, whereas the median time to recurrence in patients who died from tumor recurrence was 1.5 years (P = 0.03). Median survival was 6.2 years in patients alive with recurrence versus 1.9 years in those who died as a result of tumor recurrence.
AFP levels were available for 26 patients. The mean AFP level was 194 ± 810 ng/mL prior to transplantation with a range of 3 to 3807 ng/mL. Survival analysis based on AFP grouping at a level of 300 ng/mL did not reach statistical significance (Fig. 4).
Comparison of LDALT to Deceased Donor Liver Transplantation (DDLT)
During the same time period, 74 patients underwent DDLT for HCC at our institution. Comparing the 2 groups, we found no difference in the mean age at transplant, AFP level, largest pretransplant tumor size, pretransplant number of tumors, tumor size or number of tumors on the explanted liver, MELD score, or time to recurrence. The 3- and 5-year survival rates for DDLT were 70% and 58%, respectively (Fig. 5). In calculating the rate of tumor recurrence, we excluded 9 patients in the DDLT group because of death within 3 months following transplantation. The deaths were secondary to sepsis (2), vascular thrombosis (2), intraoperative death (1), primary graft nonfunction (1), hemorrhagic pancreatitis (1), ruptured splenic artery aneurysm (1), and respiratory failure (1). Tumor recurrence occurred in 7 patients (12.1%), and this was significantly less than the recurrence in the LDALT group (P < 0.05).
Our results for LDALT for HCC demonstrate overall survival rates at 1 and 5 years of 96% and 81%, respectively. These survival rates are comparable to those achieved in other centers with extensive experience with LDALT.4, 8–10 The majority of the patients in our study had tumors within the Milan criteria. This group of patients had a 5-year survival rate of 87.1% versus a 5-year survival rate of 80% in patients with tumors beyond the Milan criteria but within the UCSF criteria. A large multicenter study of 237 patients in Korea who underwent LDALT for HCC showed a 3-year survival rate of 91.4% for those within the Milan criteria and a 3-year survival rate of 90.6% for those within the UCSF criteria.4 On the basis of the acceptable survival of patients with tumors beyond the Milan criteria and within the UCSF criteria, some have argued for the expansion of the current UNOS policy for providing tumor exception points.9, 11 Our observed 5-year survival rate of 80% in patients with tumors beyond the Milan criteria supports the idea that the current UNOS criteria for transplantation for HCC tumors may be too restrictive and that there are select patients with tumors beyond the Milan criteria that will benefit from transplantation. The number of patients in our study was too low for us to formulate this conclusion independently. Our findings are merely in accord with other previously published studies.
The limitation of the current UNOS criteria is that they are based on the tumor number and size as determined on pretransplant imaging.12 The premise is that the tumor size and number are predictors of biological behavior. Many studies have examined the predictive factors for HCC recurrence following liver transplantation. Findings have not been uniform across studies, but some of the most common predictive factors for tumor recurrence have been the histological grade, AFP level, tumor size > 5 cm, bilobar distribution, and vascular invasion.3, 4, 13, 14 In our study, we found the histological grade to be a significant determinant of recurrence and overall survival. None of the patients with poorly differentiated tumors lived beyond 2.8 years as all of them developed recurrence. None of the patients with well-differentiated tumors developed tumor recurrence. Our data support the fact that the histological grade is likely the most reliable predictor of recurrence. In the majority of cases, the histological grade is not available prior to listing for transplantation because of concerns related to tumor dissemination after percutaneous biopsy. Also, the current imaging modalities can diagnose HCC with relatively high accuracy, thus obviating tumor biopsy.
Under the current UNOS allocation system for HCC, the median time to transplantation for all patients in the United States with stage 2 HCC who received exception points was 54 days in 2005.2 Our institution is located in Organ Procurement and Transplantation Network Region 1, in which the median time to transplantation for patients with HCC in 2006 was 168 days. Patients without HCC and MELD scores of 20 to 29 had a median time to transplantation of just over 275 days in our region.2 This difference in the wait time to transplantation has rendered LDALT a major therapeutic modality for transplant candidates who will otherwise have a higher wait-list mortality rate because of the longer waiting periods. Centers in Japan and Korea have reported substantial experience with living donor liver transplantation for HCC, demonstrating 3-year survival rates of 69% and 73% in their patients.3, 4 In the United States, the patients in the LDALT group of the A2ALL cohort study had a 3-year survival rate of 67% and a tumor recurrence rate of 29%.5 The rates of recurrence in some studies from Asian centers after LDALT for HCC have ranged from 15% to 29%.4, 8, 15 One study found a significant difference in recurrence between patients within the Milan criteria and those beyond the Milan criteria.8 In our study, we also found a higher recurrence rate in patients beyond the Milan criteria, although it was not statistically significant likely because of the small number of patients in the group beyond the Milan criteria. The higher recurrence rate in LDALT has been partly attributed to shorter waiting periods for live donor liver grafts; this has eliminated the dropout seen on the waiting list for DDLT due to tumor progression. Another proposed mechanism for higher recurrence in LDALT patients is promotion of tumor growth by up-regulated factors during the natural course of liver regeneration of a partial liver graft.16, 17 In our study, we found a strong correlation between tumor grade and tumor recurrence. All patients with poorly differentiated tumors developed recurrence and died from their recurrence, whereas 4 patients with a lesser tumor grade and posttransplant recurrence had a median survival of over 6 years following transplantation.
Despite the higher recurrence rates in the LDALT group versus the DDLT groups, patients in our study demonstrated excellent 5-year survival. In the LDALT group, the only deaths occurred in patients with tumor recurrence, and this indicated that no patients died secondary to primary graft complication or nonfunction. This is most likely due to the very short ischemic times and careful donor and recipient selection, which result in better graft quality. The relatively high tumor recurrence rate in our LDALT group seems to be the tradeoff for a better quality graft. The main limitation of our study was the small number of patients in the LDALT group. Even though our results demonstrated statistical significance in some of our analyses, larger, multicenter studies will likely provide a greater number of patients and therefore more reliable statistical results.
Overall, from an oncological standpoint, the survival rates in our study are quite acceptable. Given the overall survival rates, LDALT should still be considered a valuable option for patients with HCC, especially in areas with a limited deceased donor pool. Of course, this should be coupled with minimal donor morbidity. Efforts to develop noninvasive methods to better predict the biological behavior of tumors after transplantation will be of paramount importance in order to minimize the rate of tumor recurrence.