PDF
Citation
Print
INTRODUCTION
Hepatocellular carcinoma (HCC) is the most common indication for living donor liver transplantation (LDLT) globally [1,2] and liver transplantation (LT) is widely considered the best treatment for HCC [3]. Our center is no exception, where HCC accounts for over 50% of all adult LDLT cases annually (Fig. 1). It is projected that HCC cases will continue to increase in the future [4,5], leading to a higher demand for LT [6,7]. Moreover, paucity of deceased donors and regional trends have also contributed to the increase in the number of LDLTs for treating HCC [7-9]. To meet this demand, a significant progress has been made in surgical techniques [10], perioperative management, and immunosuppression regimens for LT [11,12].
Various selection criteria have been proposed to expand the eligibility and enhance the outcomes of LT in HCC, based on the tumor’s morphology [13-15]. However, there are still limitations to the morphological criteria, such as high recurrence of HCC and advanced tumor stages, and further increasing demand have led to a shift in transplantation eligibility criteria from morphology to biological markers in HCC treatments [16-20].
Post-LT immune suppression regimens have evolved, and the use of mammalian target of rapamycin (mTOR) inhibitors has become more common due to the adverse effects of calcineurin inhibitors (CNIs) [21,22]. Moreover, meta-analyses and numerous studies have shown that mTOR inhibitors are effective in LT and lead to improved survival outcomes [23-26]. Additionally, our recent study revealed that using pure laparoscopic donor hepatectomy led to better outcomes in 556 patients [10].
Therefore, we hypothesized that changes in selection criteria, surgical methods and post-operative management have led to better outcomes over the years. However, there is still a lack of long-term outcome data on LDLT in HCC, including a decade of follow-up, a large number of patients, and real-world evidence. Therefore, we aimed to evaluate the differences in outcomes between two distinct periods of HCC LDLT cohorts. The earlier decade focused on morphological criteria, while the recent decade prioritized additional biological tumor markers in HCC LDLT. Our primary measures include overall survival (OS), recurrence-free survival (RFS), and mortality rate.
MATERIALS AND METHODS
A retrospective analysis was conducted on patients who were diagnosed with HCC and underwent LDLT at Seoul National University Hospital (SNUH) between January 2000 and June 2022. The study population was divided into two groups: Group A (January 1, 2000 to June 30, 2011) and Group B (July 1, 2011 to June 30, 2022). Patients with intra-hepatic cholangiocarcinoma (IHC) or combined IHC-HCC were excluded. The latest follow-up date for the study was January 1, 2024.
The study inclusion criteria were as follows: (i) age ≥18 years with written consent, (ii) histologically proven HCC with a degree of necrosis from 0 to 99 percent, (iii) unviable tumor or total necrotic nodules with a prior history of surgical resection resulting in HCC pathological confirmation, (iv) unviable tumor or total necrotic nodules with a prior history of locoregional therapies (LRT) such as transarterial chemoembolization, radio frequency ablation, and percutaneous ethanol injection therapy (PEIT).
The baseline characteristics, such as age, sex, and underlying disease, were reviewed. Preoperative model for end-stage liver disease (MELD) score, Child-Turcot-Pugh class, alpha-fetoprotein (AFP), protein-induced vitamin K absence (PIVKA-II), and prior history of surgical or LRT were investigated. LRT was defined based on type of interventions either single method or combined methods. Transarterial radioembolization was not included. Perioperative total operation time and graft liver volume were also analyzed. The graft-recipient-weight ratio (GRWR) calculation was based on the post-surgical graft volume (grams).
The Milan criteria were determined based on the extent of explant pathology reports. Unviable tumors or total necrotic nodules were considered within the Milan criteria. Far-advanced HCC was defined by tumor numbers greater than 10 or larger than 10 cm or the presence of macrovascular invasion. In explant histologic variables, explant tumor differentiation was assessed according to either WHO tumor differentiation grade (well to undifferentiated) or Edmondson-Steiner’s grade (I–IV). The largest viable tumor size, number of total viable tumors, and extent of microvascular invasion were also scrutinized. The OS rates and RFSs were shown at 1, 3, and 5 years, as there was a difference in the total follow-up periods between the two different time points.
Statistical Analysis
Statistical analysis was performed using SPSS software version 29 (IBM Corp., Armonk, NY, USA). Student’s t-test or Kruskal–Wallis’s test were used on the continuous variables and results are shown as a median value with an interquartile range (IQR). Chi-square test and Fisher’s exact tests were used on the categorical variables, and results are shown as observed numbers with percentages. OS and RFS curves were calculated using Kaplan–Meier method, and comparisons were made using a Log-rank test. A p-value less than 0.05 was considered significant.
RESULTS
A total of 942 patients were included in this study. Table 1. presents the baseline characteristics of the study population, divided into Group A (n=314, 2000.01.01–2011.06.30) and Group B (n=628, 2011.07.01–2022.06.30). The median age was similar in both groups, with Group A at 56.0 (IQR 52.0–63.0) and Group B at 57.0 (IQR 52.0–62.0) (p=0.85). Males accounted for 83.4% of Group A and 80.6% of Group B (p=0.32).
Regarding underlying disease, a larger percentage of both Group A and Group B were affected by hepatitis B virus Infection (77.7% vs. 74.4% in Group B, Table 1, p<0.01). The number of resections and LRT has significantly increased in recent periods (Group B). Moreover, in Group B, preoperative liver resection was performed on more patients (17.7% vs. 2.9% in Group A, p<0.01).
Tumor Biological Markers and Locoregional Therapies
We assessed the liver function and disease status of patients prior to surgery by utilizing LRT status, the MELD score, and the Child-Turcot-Pugh score. Group B had a considerably higher percentage of patients in Class A (62.4%) than Group A (37.2%), signifying a significant difference (p<0.01). The median MELD Score was significantly higher in Group A (17.4 [IQR 12.7–23.3]) compared to Group B (9.0 [IQR 7.2–12.9]), with a p-value of <0.01.
Single type LRT like trans-arterial chemoembolization (TACE) or radiofrequency ablation (RFA) or PEIT were undergone by 55.6% of patients in Group B, which was significantly higher than the 44.9% in Group A (p<0.01). Double LRT, such as TACE with RFA or PEIT, were more commonly used in Group B at 25.2% compared to 17.8% in Group A (p<0.01). The usage of Triple methods, TACE+RFA+PEIT, was also higher in Group B (4.9%) compared to Group A (2.5%), with a significant difference (p<0.01). Importantly, the proportion of patients who didn’t receive any LRT was significantly lower in Group B (14.3%) compared to Group A (34.7%), with a p-value of <0.01, indicating a statistically significant increase in the use of LRT in Group B.
Regarding the tumor’s biological characteristics before surgery, Group A had a significantly higher median AFP level of 15.0 ng/mL (IQR 5.5–84.5) than Group B’s median of 5.4 ng/mL (IQR 2.9–16.6), with a p-value of less than 0.01. However, there were no significant differences in PIVKA-II levels. Group A recorded a median level of 30.0 mAU/mL (IQR 16.0–90.5) compared to Group B’s 27.0 mAU/mL (IQR 18.0–67.0) (p=0.31). Before the surgery, Group B showed significant improvement in liver function and a decrease in preoperative AFP levels. Based on these findings, it can be concluded that Group B had a better preoperative status compared to Group A.
According to the Milan criteria, the analysis of the explant pathology showed that 72.6% of patients in Group A met the criteria, which was slightly lower than the 75.5% in Group B (p=0.50). The percentage of patients who were classified as far advanced and beyond the Milan criteria was similar in both groups, with minor differences between them. In Group A, 21.3% and 6.1% of patients were classified as far advanced and beyond the Milan criteria, respectively, while Group B had 18.2% and 6.4% in the same categories, respectively (p=0.50). Fig. 2 shows the annual proportion of cases that exceed the Milan criteria, including those that are beyond Milan and Far advanced at SNUH.
Moreover, there was a significant difference in operation times, with Group A having a longer median time of 470.0 minutes (IQR 420.0–540.0) compared to 385.0 minutes in Group B (IQR 330.0–454.7; p<0.01). Graft liver volume was slightly higher in Group A, with a median of 715.0 mL (IQR 645.0–823.5), vs. 702.0 mL in Group B (IQR 612.0–803.0; p=0.03). The GRWR also showed a significant difference, with Group A having a higher median GRWR of 1.16 (IQR 0.97–1.16) compared to 1.05 in Group B (IQR 0.89–1.05; p<0.01). Tumor differentiation showed significant differences, with Group A showing poorer differentiation. Total necrosis was significantly higher in Group B (28.0% vs. 13.4%; p<0.01). Group A had larger tumors compared to Group B (2.4 cm vs. 2.2 cm; p=0.01) (Table 2).
Recurrence and Survival Rates
Both Group A and Group B were compared in terms of post-surgical RFS, and the results were as follows: at 1 year, Group B had a significantly higher RFS rate of 90.6% compared to Group A’s rate of 86.3% (p=0.03) (Fig. 3A). However, after 3 years, both groups experienced a decline in their RFS rates to 84.4% and 80.0% for Group B and Group A, respectively. Finally, at the 5-year mark, the RFS rate for Group B was slightly higher than that of Group A at 82.8% and 78%, respectively.
Additionally, Group B exhibited better OS rates than Group A. Specifically, Group B demonstrated a 1-year OS rate of 95.9%, whereas Group A had a rate of 91.7%. At 3 years, Group B’s OS rate was 87.9%, and Group A’s was 80.6%. Finally, at the 5-year mark, Group B had an OS rate of 84.6%, which was significantly higher than Group A’s rate of 77.4% (p<0.01) (Fig. 3B).
Furthermore, the recurrence rate was higher in Group A (23.6%) compared to Group B (17.0%, p=0.02), but there was no significant difference in the median time to recurrence between the two groups. The median time to recurrence was 9.1 months (IQR 3.9–21.8) for Group A and 11.4 months (IQR 6.6–18.5) for Group B (p=0.92). Moreover, Group B had a longer median tumor-bearing survival time (20.0 months [IQR 5.4–25.7]) compared to Group A’s median tumor-bearing survival time (12.3 months [IQR 5.2–26.1], p<0.01). Additionally, the overall mortality rate was significantly higher in Group A (34.7%) compared to Group B (16.1%, p<0.01).
In summary, Group B had better OS rates, RFS rates, and median tumor-bearing survival time than Group A. Conversely, Group A had a higher recurrence rate and overall mortality rate.
DISCUSSION
Biological Tumor Markers Refined Selection and Provided Better Survival Results
Our study compared the long-term results of LDLT in HCC between two time periods. We observed a significant decrease in the median MELD score (Group A: 17.4, Group B: 9.0, p=0.01) and AFP levels (Group A: 15.0 ng/mL, Group B: 5.4 ng/mL, p<0.01) from the past to the recent period. The higher median AFP level in Group A may indicate more aggressive or advanced disease before transplantation. On the other hand, the lower AFP in the recent cohort can be attributed to the improved selection process for LT candidates, based on their tumor characteristics (e.g., AFP and PIVKA-II). Previous studies have suggested that a low AFP can lead to better survival and serve as an important prognostic predictor for HCC recurrence [27,28]. Our study also validates this hypothesis, as Group B, representing the most recent decade, achieved superior outcomes in RFS and 1-to-5-year OS rates of 95.9% and 84.6%, p<0.01. Although median time to recurrence did not show statistical significance, Group B had slightly longer median time to recurrence, indicating improved outcomes. Additionally, the overall mortality rate decreased in Group B (Group A: 34.7%, Group B: 16.1%, p<0.01). This may be related to the lower RFS and recurrence rate in Group B (Group A: 22.6%, Group B: 17.0%, p=0.02), as high tumor recurrence is associated with high mortality rate and recurrence is the major cause of post-surgical mortality.
Standardized Protocols and State-of-the-Art Technologies Reduced Operation Time
At our center, every stage of the LT process follows a standardized routine protocol, and most living donor hepatectomies have been performed through pure 3D laparoscopy since 2015 [10]. These factors have likely contributed to the decreased total operation time and improved surgical outcomes in Group B, along with decreased cold and warm ischemic times for the graft liver. The implementation of standardized protocols and advancements in laparoscopic surgery technology have reduced operation time and improved outcomes in recent Group B.
Locoregional Therapies were Extensively Utilized Recent Years and Yielded Better Results
The proportion of LRT was higher in Group B, where most patients received a single method of TACE, RFA, or PEIT (Group A: 44.9%, Group B: 55.6%). No LRT received patients were higher in Group A (Table 1). These results suggest that the majority of patients in Group B either exceeded the Milan criteria or were ineligible for surgery, thus requiring downstaging therapy to reduce tumor burden or bridging therapy for LT. As a result of these downstaging and bridging therapies, analysis of the tumors upon removal showed that more than 70% of patients in both groups met the Milan criteria (Table 2). Moreover, the percentage of completely necrotic nodules was higher in Group B (28.0%) compared to Group A (13.4%), indicating a greater use of LRT and a more successful therapeutic effect of LRT in Group B. This ultimately contributed to a higher survival rate and a lower recurrence rate in Group B.
Use of mTOR Inhibitors Improved Outcomes
CNIs, such as tacrolimus, have long been established as the standard immune suppression regimen after organ transplantation. However, CNIs have a limited range of therapeutic effectiveness: inadequate levels can lead to acute rejection, while excessive levels can result in toxic effects associated with CNIs [21,22]. Because of narrow trough level and adverse effects, there has been a shift in the use of immune suppression drugs. Since 2014, mTOR inhibitor drugs, such as sirolimus and everolimus, have been widely accepted as an alternative long-term (over 3 months) immunosuppression regimen. These drugs have demonstrated tumor suppressive effects and protection for the kidneys, resulting in improved long-term survival outcomes for liver transplant patients [25]. With this change, our center has regularly implemented mTOR inhibitors for advanced HCCs [29]. Specifically, in Group B, mTOR inhibitors were more commonly used as maintenance therapy and long-term immune suppression one month after transplantation, aiming to achieve clinical tolerance. Therefore, it is reasonable to assume that the increased use of mTOR inhibitors as a maintenance regimen in the recent decade has resulted in improved survival rates.
There are limitations to our study, as is typical of retrospective studies. Our analysis relied on the completeness of the medical records, which may introduce potential biases. Additionally, our findings may not be applicable to patients who underwent deceased donor liver transplantation (DDLT) as our study only included patients who underwent LDLT, not DDLT.
Nevertheless, our findings suggest that Group B, recent period, demonstrated superior survial results and decreased mortality rates than Group A, past period. Enhanced use of mTOR inhibitors and improved patient selection criteria, particularly focusing on tumor biological characteristics, have contributed significantly to these chronological improvements. Further pervasive use of biological tumor marker based patient selection criteria should be encouraged.
XML Download