Objectives: Our aims were to determine whether transarterial chemoembolization before liver transplant for hepatocellular carcinoma improves posttransplant survival and whether patients downstaged by transarterial chemoembolization within Milan criteria have a posttransplant survival benefit.
Materials and Methods: In this retrospective analysis of prospectively collected data, survival rates of 87 patients treated with and 68 patients not treated with transarterial chemoembolization before liver transplant were compared using 2-sample t tests and multivariate Cox regression. We also compared posttransplant survival of patients within Milan criteria versus those downstaged after transarterial chemoembolization. We controlled for disease severity by assessing, among other variables, tumor diameter before and at transplant and alpha-fetoprotein levels before transplant and transarterial chemoembolization.
Results: Overall 1-, 3-, and 5-year survival rates were 84%, 71%, and 63%, respectively. These rates were 91%, 78%, and 73% for patients who received and 76%, 63%, and 54% for patients who did not receive transarterial chemoembolization. Hazard ratios were 0.56 for having versus not having transarterial chemoembolization (P = .04), 1.06 for total tumor diameter on explantation (P = .01), 1.5 for largest tumor > 3 cm (P = .15), and 2.9 for pretransplant alpha-fetoprotein > 659 ng/mL (P = .006). A higher end-stage liver disease score correlated with poorer overall survival (hazard ratio = 1.53; P < .001). Laboratory values, lipiodol uptake, imaging response, and downstaging into Milan criteria were not correlated with survival.
Conclusions: Patients with hepatocellular carcinoma who were treated with neoadjuvant transarterial chemoembolization had better survival rates posttransplant than those not treated with transarterial chemoembolization. A high pretransplant alpha-fetoprotein level was negatively correlated with survival. Patients downstaged to Milan criteria after transarterial chemoembolization fared equally well versus those who met Milan criteria initially. Pretreatment with transarterial chemoembolization was positively correlated with survival posttransplant, with patients having a 44% reduction in posttransplant mortality.
Key words : Alpha-fetoprotein, Chemoembolization, Hepatocellular carcinoma, Locoregional therapy, Milan criteria
The adoption of the Milan criteria1 and the San Francisco criteria2 for orthotopic liver transplant in patients with hepatocellular carcinoma (HCC) has resulted in favorable 5-year posttransplant survival rates of 70% to 80%.2-4 Consequently, the percentage of orthotopic liver transplants performed for cirrhosis-related HCC has increased from 7% to 22%.3,4 The average wait time for transplant based on the Model for End-Stage Liver Disease (MELD) allocation system is currently 8.3 months.4 Although this is an improvement from the 28-month wait time in the pre-MELD era, the risk of a patient dropping off the wait list during the wait period remains high, at 22%.4 Additionally, lengths of stays on wait lists continue to increase in the United States and internationally.
Transarterial chemoembolization (TACE) is the standard of care in the treatment of unresectable HCC, affording a significant survival benefit and local tumor control.5-7 Transarterial chemoembolization is widely used in transplant centers as a means of downstaging patients with unresectable HCC, thereby “bridging” patients on wait lists to surgery and preventing drop-offs.8-11 Various studies have explored the benefits of preoperative chemoembolization on posttransplant survival. However, the results are mixed, with some studies showing improved survival12,13 but most showing no differences in survival between patients who are treated with TACE before transplant and those who are not.14-21
The primary objective of this study was to ascertain whether patients with HCC who were pretreated with TACE had better survival rates than patients who were not pretreated with TACE. Secondary objectives were (1) to evaluate intergroup survival differences using imaging, laboratory results, tumor criteria, and other factors; (2) to investigate whether patients who were downstaged into Milan criteria faired as well as those within Milan criteria initially; and (3) to determine the pretransplant alpha-fetoprotein (AFP) level for which the P value reaches .05 for posttransplant survival.
Materials and Methods
This was an institutional review board-approved, retrospective cohort study of prospectively collected data. It included all patients with cirrhosis and HCC who underwent liver transplant from February 2, 2005, through July 10, 2015, at our tertiary care hospital.
A team consisting of hepatologists, oncologists, interventional radiologists, and transplant surgeons assessed all transplant-eligible patients at our institution for their eligibility for TACE during a weekly multidisciplinary board meeting. Transarterial chemoembolization protocols used were either conventional TACE using doxorubicin (50 mg), mitomycin C (10 mg), lipiodol (10 mL), and varying amounts of 100- to 300-μm Embospheres (Biosphere Medical, Rockland, MA, USA) or TACE with drug-eluting beads, using 50 to 100 mg of doxorubicin loaded onto 100- to 300-μm spheres. The choice was operator dependent, and no survival differences have been demonstrated between these treatments.22 Patients were treated with TACE if they did not meet the Milan criteria or if their HCC was within Milan criteria and larger than 2 cm. The 2-cm size was selected so that HCC patients would receive tumor points for transplantation. Patients with small HCC tumors (< 2 cm) and/or high likelihood of expeditious transplant did not undergo TACE.
The following data were captured: patient demographic characteristics (age, race, sex); underlying causes of cirrhosis; hepatitis C virus status and genotype; histologic stage of hepatic fibrosis; Milan criteria status; MELD score; ABO blood group type; laboratory test values (albumin, alanine aminotransferase, alkaline phosphate, aspartate aminotransferase, blood urine nitrogen, creatinine, hemoglobin, international normalized ratio, lymphocyte count, mean corpuscular volume, neutrophil count, platelet count, prothrombin time, total bilirubin, total protein, and white blood cell count); tumor characteristics (number of tumors, largest tumor diameter, total tumor diameter [calculated as the sum of the largest diameters of all tumors]); pretransplant TACE procedures (number, type, and location); AFP values before and after TACE and before and after transplant; percent lipiodol uptake; European Association for the Study of the Liver (EASL) scores before and after TACE (determined by reviewing imaging before and after TACE); date of liver transplant; tumor differentiation, angiolymphatic invasion, and percent tumor necrosis on explantation; and tumor recurrence, time to recurrence, patient death, date of death, and cause of death.
Angiographic and cross-sectional images were obtained from a picture archiving and communication system. Response to TACE was assessed using the EASL criteria in which 1 = 0% to 25% enhancement, 2 = 26% to 50% enhancement, 3 = 51% to 75% enhancement, and 4 = 76% to 100% enhancement. Degree of hepatic fibrosis was assessed on a 5-point scale according to the METAVIR scoring system in which F0 = no fibrosis and F4 = cirrhosis.23
Survival time was defined as the time from transplant to last follow-up or patient death. Survival data were truncated at 5 years after transplant for Kaplan-Meier analyses, but all survival data were included in regression analyses. Univariate Cox regression was used to analyze the following: whether a patient received any TACE procedures, whether the largest tumor was > 3 cm in diameter, whether the pretransplant AFP value was > 659 ng/mL (calculated AFP that minimized P-value), and whether a patient was downstaged into Milan criteria. Total tumor diameter, degree of lipiodol uptake, MELD score, and change in EASL score after receiving TACE were included in this analysis. Multivariate Cox regression incorporated the following variables found to be significant on univariate analysis: receipt of TACE, presence of a tumor > 3 cm, pretransplant AFP > 659 ng/mL, total tumor diameter, and MELD score. The inclusion of total tumor diameter and pretransplant AFP allowed us to control for disease severity between patients who did and those who did not receive TACE. The AFP values were dichotomized with a threshold that maximized the Akaike information criterion in univariate Cox regression analyses.
Significance levels (alpha = .05) were determined by comparing the estimates with the distribution obtained through permutation of the survival information across patients.
During the defined study period, 155 patients (37 women) underwent liver transplant for HCC at our institution. Median follow-up time after transplant was 2.7 years (range, 0-10 y). Median patient age was 59 years (interquartile range, 55-63 y). Ninety-four patients (61%) were white, 45 (29%) were black, and 16 (10%) were Asian. The causes of cirrhosis were hepatitis C virus (n = 103), hepatitis B virus (n = 14), alcoholic cirrhosis (n = 8), nonalcoholic steatohepatitis (n = 6), primary biliary cirrhosis (n = 1), mixed disorders (eg, hepatitis C virus, hepatitis B virus, alcoholic cirrhosis, human immunodeficiency virus, hemochromatosis) (n = 18), and other (n = 5).
Of the 155 patients, 136 (88%) met the Milan criteria at the time of presentation, and 19 (12%) were downstaged to the Milan criteria before transplant. On explantation, only 47 patients (30%) met the Milan criteria on the basis of pathologic evaluation.
Transarterial chemoembolization treatment
Of 155 transplant recipients, 87 (56%) were treated with TACE before transplant. The remaining 68 patients (44%) were not treated with TACE because of imminent transplant or small (< 2 cm) and stable lesions. The mean (standard deviation [SD]) number of TACE procedures per patient was 1.34 (0.61) (range, 1-4). Of 87 patients with TACE treatment, 44 patients (51%) underwent conventional TACE. The remaining 43 patients (49%) underwent TACE using drug-eluting beads. Of the patients who underwent TACE, the mean (SD) EASL scores before and after treatment were 3.87 (0.34) and 1.58 (0.8), respectively. Mean (SD) lipiodol uptake after all TACE procedures was 82% (25%).
On explantation, 151 livers (97%) were histologically classified as F4 (fibrosis with cirrhosis), and 4 (2.6%) were classified as F3 (septal fibrosis without cirrhosis) on the METAVIR score. Five patients (3%) were noted to have innumerable tumors throughout the liver despite meeting the Milan criteria according to preoperative imaging. For the other 150 patients, the mean (SD) number of tumors on explantation was 1.95 (1.36). The mean (SD) diameter of the largest tumors was 3.0 (1.6) cm. Mean (SD) aggregate tumor diameter was 4.8 (4.2) cm. On histologic evaluation, 24 patients (15%) had well-differentiated cancer histology, 94 patients (61%) had moderately differentiated cancer histology, 15 patients (10%) had poorly differentiated histology, and 22 patients (14%) were not classified. Forty-one patients (26%) had angiolymphatic invasion on explantation.
Twenty-three patients (15%) had tumor burden that was noted to be nonviable on pathologic assessment after liver explantation. The remaining 132 patients (85%) had varying degrees of tumor viability, with tumor necrosis ranging from 0% to 95%. The mean (SD) MELD score before transplant was 7.5 (1.1). Multiple laboratory values were obtained pretransplant (Table 1).
Tumor recurrence and survival
Tumors recurred in 22 of 154 patients (14%) (data on tumor recurrence were unavailable for 1 patient). Among patients with recurrence, the median time to recurrence was 314 days (interquartile range, 93-557 days). At the time of data analyses, 132 patients (86%) were free from recurrence.
At the time of data analyses, 54 patients had died. Causes of death included disease recurrence (n = 19), sepsis (n = 18), liver failure (n = 7), cardiac disease (n = 5), renal failure (n = 2), cerebrovascular accident (n = 1), lung cancer (n = 1), and pulmonary embolism (n = 1).
Table 2 shows the 1-, 3-, and 5-year survival rates. Patients in the TACE group had better survival rates at all 3 periods compared with patients in the non-TACE group (P = .04; Figure 1). Because the hazards were parallel for the first 5 years, there is only 1 P value for the whole period.
On univariate analysis, the hazard ratio (HR) for TACE versus non-TACE was 0.56 (95% confidence interval [CI], 0.32-0.98; P = .04). For total tumor diameter on explantation, HR was 1.06 (95% CI, 1.01-1.12; P = .01); for largest tumor > 3 cm, HR was 1.5 (95% CI, 0.87-2.59; P = .1); and for pretransplant AFP > 659 ng/mL, HR was 2.9 (95% CI, 1.36-6.22; P = 0.006). Higher MELD score was correlated with poorer survival, with HR of 1.53 (95% CI, 1.21-1.92; P < .001). Laboratory values, lipiodol uptake, EASL imaging response, and downstaging into Milan criteria were not correlated with survival.
On multivariate analysis, the adjusted HR for TACE versus non-TACE was 0.73 (95% CI, 0.40-1.34; P = .3). For total tumor diameter on explantation, the adjusted HR was 1.02 (95% CI, 0.96-1.09; P = .05); for largest tumor > 3 cm, adjusted HR was 1.58 (95% CI, 0.84-2.96; P = .2); for pretransplant AFP > 659 ng/mL, adjusted HR was 2.76 (95% CI, 1.24-6.13; P = .01); and for MELD, adjusted HR was 1.51 (95% CI, 1.15-1.98; P = .01).
The mean (SD) values for AFP were 390 (2670) ng/mL before TACE, 251 (1808) ng/mL after TACE, 390 (2670) ng/mL immediately before transplant, and 15 (68) ng/mL immediately after transplant.
Approximately 22% of patients with HCC drop off the liver transplant wait list. In half of these patients, this is due to tumor progression.4 Frangakis and associates8 showed that chemoembolization is associated with a significant decrease in the risk of patients with HCC dropping off the wait list. As wait times for liver transplant grow and the need for organ procurement becomes more acute because of the increasing incidence of obesity-related liver disease, the role of TACE in maintaining patients with HCC on wait lists will become more significant. Without chemoembolization, many of these patients will not be downstaged into Milan criteria, and those who already meet the Milan criteria will be at greater risk for dropping off the list. In our study, 12% of patients who received a liver transplant did not initially meet the Milan criteria but were downstaged with TACE before transplant. Among patients who received TACE, the 1-, 3-, and 5-year survival rates were 91%, 78%, and 73%, respectively, versus 76%, 63%, and 54%, respectively, in patients without TACE (P = .04).
Our results are in accordance with several other studies that noted a survival benefit in patients receiving locoregional therapy (LRT) before liver transplant. Bharat and colleagues12 performed a retrospective study of 100 patients who underwent orthotopic liver transplant, 46 of whom received some form of LRT (radiofrequency ablation, TACE, ethanol injection, or a combination thereof) before transplant. Thirty-six patients (78%) received LRT with TACE, whereas 4 patients (9%) underwent TACE in combination with ablation. Of the patients who received LRT, the 5-year survival rate was 82% versus 52% for patients who did not receive LRT (P = .01). Although the LRT group included patients who also underwent radiofrequency ablation and ethanol ablation, these modalities were used in a minority of patients. Yao and associates2 showed that pretransplant LRT in patients with pathologic T2 or T3 HCC conferred a significant survival benefit. Seventy-three patients (71%) in their LRT group underwent TACE only, whereas another 15 patients (15%) underwent TACE in combination with ablation. The 1-year and 5-year recurrence-free survival rates, respectively, were 96.4% and 93.8% in patients who underwent LRT versus 91.5% and 81.6% in patients who did not (P = .049).
The primary limitation of our study was its retrospective nature, which is mitigated by the prospective data collection.
In conclusion, we found that neoadjuvant TACE was associated with longer overall survival in patients who received a liver transplant for HCC. In addition, patients downstaged into Milan criteria with TACE appeared to experience a posttransplant survival rate similar to that of patients who met the Milan criteria initially.
DOI : 10.6002/ect.2018.0017
From the 1Division of Vascular and Interventional Radiology,
Department of Radiology, The Johns Hopkins University School of Medicine; the
2Bloomberg School of Public Health, The Johns Hopkins University; the
3The Johns Hopkins School of Medicine; the 4Department of
Surgery, The Johns Hopkins University; the 5Division of
Transplantation, Department of Surgery, The Johns Hopkins University; and the
6Division of Gastroenterology and Hepatology, Department of Medicine,
The Johns Hopkins University, Baltimore, Maryland, USA
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Christos Georgiades, Division of Vascular and Interventional Radiology, The Johns Hopkins University School of Medicine, 1800 Orleans Street, Baltimore, MD 21287, USA
Phone: +1 410 614 1046
Table 1. Laboratory Values Obtained Before Liver Transplant for 155 Patients With Hepatocellular Carcinoma
Table 2. Survival Rates After Liver Transplant by Transarterial Chemoembolization Treatment Status for Hepatocellular Carcinoma
Figure 1. Kaplan-Meier Estimates Showing Overall Survival Rate of 155 Patients, Rate for 87 Patients Who Received Transarterial Chemoembolization Before Liver Transplant, and Rate for 68 Patients Who Did Not Receive Transarterial Chemoembolization Before Liver Transplant