Non-Resective Ablation and Liver Transplantation in Patients with Cirrhosis and Hepatocellular Carcinoma (HCC): Safety and Efficacy
Corresponding author: Robert A. Fisher, rafisher@hsc. vcu.edu
We investigated the efficacy of nonresective ablation techniques and the tumor-free survival of cirrhotic patients undergoing liver transplantation for hepatocellular carcinoma (HCC). In group 1, 11 HCC patients were treated with these techniques and transplanted. On the waiting list, patients were treated to complete ablation, judged by gadolinium-enhanced MRI and alpha-fetoprotein (AFP) levels. Group 1 was compared with a concurrent group of 10 liver transplant patients (group 2) with incidental HCC (stages T1 three patients, T2 seven patients). The group 1 patients received 36 procedures (4 alcohol ablations, 14 trans -hepatic artery chemo-embolizations, 15 trans -hepatic chemo-infusions, and 3 radio frequency ablations) for treatment of 13 liver masses. Tumor-node-metastasis (TNM) stage was reduced in eight patients (72.7%), unchanged in two patients and increased in one patient before transplantation. The mean waiting time for transplantation was 12.9 7.6 months. Both groups had a tumor-free survival of 100%, at 30 12 months post transplant. On pathology, 54.5% of explanted livers had residual viable HCC after tumor treatment, and 36.4% (4/11) explants had synchronous lesions. Non-resective ablation therapy is safe and effective in reducing the HCC progression in cirrhotic patients awaiting liver transplantation. The cancer-free survival rate in this treatment group is equal to that for incidental T1–T2 HCCs.
Hepatocellular carcinoma (HCC) is the most common primary malignant tumor of the liver and one of the major causes of death among patients with cirrhosis (1). Underlying cirrhosis is found in more than 85% of patients with HCC (2,3).
In the United States, with the explosion of chronic liver disease from viral hepatitis in adults, and the well-established association of malignant degeneration in the cirrhotic liver, the increasing incidence of hepatocellular carcinoma (HCC) is a sobering reality (4). Knowing that most patients in the US with the diagnosis of HCC have moderate-to-severe liver disease and cirrhosis, optimum care requires the complex analysis of cancer stage, to predict recurrence, and determination of liver reserve to predict the suitability of resection vs. total hepatic replacement to prevent death from liver failure (3–7).
Awareness of the potential for cirrhosis-associated hepatocellular carcinoma has led to the establishment of screening programs, based on ultrasonography and the measurement of α-fetoprotein levels, in high-risk populations in eastern countries and in patients awaiting liver transplant in western countries (7). These screening programs have been associated with earlier detection of small tumors, and with less severe liver disease at the time of diagnosis (8). With HCC lesions identified earlier in an increasing number of patients, along with the development of new therapeutic techniques, such as local-disease control techniques and liver transplantation, the approach to the treatment of HCC in cirrhotic patients has changed in the last decade.
Surgical resections for the treatment of HCC have been used for years and remain a mainstay of treatment of HCC in patients with early-stage cirrhosis, but a high HCC recurrence rate has been reported (9). The multifocal nature of HCC in cirrhotic patients explains the high rate of recurrence after treatment with resection or with any local-control technique (10). Schwartz and associates, at the Mount Sinai Medical Center, showed multifocality in 38% of explanted livers of transplant recipients with HCC less than 2 cm in diameter and in 82% of patients with HCC tumors smaller than 5 cm in diameter (11).
Non-surgical methods for local-disease control, such as percutaneous ethanol injection (PEI), radiofrequency thermal ablation (RFA), trans-hepatic artery chemoembolization (TACE), trans-hepatic artery chemoinfusion (TACI) and laser ablation (LA) have evolved in the last 20 years as emerging technology for the treatment of liver tumors.
In the late 1960s, the use of liver transplantation in unselected patients with HCC was discouraging. Many authors reported survivals between 15 and 35% at 5 years, using liver transplantation without adjuvant chemotherapy (12–14).
Nowadays, liver transplantation has become widely accepted as therapy for hepatocellular carcinoma when liver resection is not possible, or for small (stage I–II) HCC. The combination of local-disease control techniques and liver transplantation for selected group of patients with HCC (T1–T2 stages) seems to have encouraging results (10,15–19). This combined approach appears to be the appropriate answer for both the cancer and the primary liver disease (19).
This study investigates the efficacy of nonresective ablation techniques and the tumor-free survival of cirrhotic patients undergoing liver transplantation for HCC. A prospective pathology review of all explanted livers addresses the ‘cancer kill’ histologic efficacy of the nonresective ablation techniques and accuracy of sophisticated imaging techniques in ‘seeing’ viable HCC.
Patients and Methods
From December 1997 to July 2001, 70 patients with hepatocellular carcinoma were evaluated from a total of 508 patients referred for orthotopic liver transplantation (OLT) evaluation and/or resection by the Transplant Division of the Virginia Commonwealth University, Medical College of Virginia Hospitals. Five patients, after being informed of the advanced nature of their cancer and/or medical disease, chose not to participate. Seventy patients underwent complete history and physical exams; blood testing for hematologic profile, hepatic panel [T. bilirubin mg/dL, albumin g/dL, prothrombin time (PT) s, INR], serologic study (anti-HCV, HEPB surface Ag, HB core Ag), serum creatinine mg/dL, alpha-fetoprotein ng/mL; classification of cirrhosis, and Child-Turcotte Pugh (CTP) scoring of severity of liver disease.
The initial radiologic staging was based on ultrasonographic screening of the liver and vasculature, followed by half-Fourier rapid acquisition with relaxation enhancement (RARE) magnetic resonance imaging angiography (MRI/MRA) of the liver, contrast-enhanced CT of the chest and head, and total body bone scan (20,21). All questionable HCC diagnoses were confirmed by biopsy, and three patients not included in the 65 HCC patients were initially screened for HCC but found to have other tumors, i.e. dysplastic nodule and focal nodular hyperplasia, and not HCC. All HCC patients were staged initially and serially with the American Liver Tumor Study Group modified tumor-node-metastasis (TNM) staging classification (22). Timing of the follow-up study restarted with the Q 3 month timing again from date of transplant, resection and complete HCC ablation.
In this study, a subgroup of 11 selected patients (group 1) received multimodality therapy before being enrolled on the waiting list for liver transplantation. All patients in this group were followed with monthly AFP, and half-Fourier RARE MRI of the liver, chest CT scan and total body bone scan every 3 months, as well as follow-up in the transplant clinic every 3 months. Morbidity, mortality and cancer-free survival were evaluated.
Group 1 was compared with a group of 10 cirrhotic patients (group 2) undergoing liver transplantation because of hepatitis C virus (HCV) cirrhosis who had the finding of HCC on their explanted livers (incidental HCC).
Four ablation techniques were utilized over the course of study to reduce tumor bulk, to prevent porto–portal and systemic micrometastases on tumor manipulation; to completely ablate HCC in the liver to prevent disease progression in patients awaiting liver transplantation for concomitant end-stage liver disease; as primary treatment for cure in patients not transplant or resection candidates, at the time of initial staging; and to palliate life-threatening complications of HCC (i.e. HCC hemorrhage, rupture, etc.). The four ablation therapies included: transcatheter hepatic arterial chemoembolization (TACE); transcatheter hepatic arterial chemoinfusion (TACI); fine-needle absolute alcohol ablation (EtOH Ab); and radiofrequency microwave ablation (RFA). Individual decisions on the safest and most effective ablation techniques to use were based on the senior author (R.A.F.)'s evaluation of HCC size, location and multiplicity; limits of imaging localization for accurate, total ablation; and estimation of the zone of thermal and chemical injury in relation to the estimated hepatocellular reserve and risk of systemic morbidity of each ablation technique. In general, ablation techniques were developed as outpatient procedures, which played a role in the design of treatment algorithms. For example, early in our experience, EtoH ablation where EtoH penetrated Glisson's capsule into the peritoneum, caused immediate and severe abdominal pain (peritonitis), leading to added hospitalizations and testing. To avoid this problem, we use TACE and or RFA for HCC abutting Glisson's capsule. If MRI-visualized HCC could not be identified for needle localization by ultrasound, then TACE was chosen over RFA. In 1998, early in the senior author's experience with RFA, tumors less than 0.5 cm from a major venous branch on MRI localization were selected for TACE to theoretically prevent an inadequate margin of ablation by RFA on the tumor side of a large vessel that would cause a ‘heat sink’.
TACE. All patients were admitted the morning of the scheduled procedure for intravenous (i.v.) crystalloid hydration, and hematologic, metabolic and hepatic panel laboratory evaluation. Two hours pre-angio, intravenous dexamethasone, ondansetron HCL, and Timentin® were administered. After angiographic catheter placement by Seldinger technique, 25 g of mannitol was infused, and cisplatin (50 mg/m2) and doxorubicin (25 mg/m2) suspended in ethiodol oil (5–8 mL) was selectively infused into the tumor/tumors, followed by tumor-selective arterial embolization with gelatin sponge (Gelfoam, Upjohn, Kalamazoo, MI) until 2000 when sponge use was replaced by 500–700-μm microspheres (Biosphere Medical Inc., Rockland, MA). Only arteries supplying the tumor were embolized, and no regional hepatic artery embolization was performed. On completion of angiographic procedures, intravenous furosemide was infused, and postangio checks were followed with additional hydration, q8 h antibiotic, anti-emetic, antacid administration, and q4 h hydromorphone administration for pain overnight. Hospitalization for >23 h, and indications, were recorded as morbidity, as was use of granulocyte colony stimulating factor (GCSF) for total WBC less than 3 × 103 cells/mm3.
TACI. Chemoinfusion was administered 2–3 weeks after ablation by TACE, EtOH or RFA. Patient care was identical to TACE, without overnight hospitalization; and chemotherapy of equal dose without Ethiodol, and without Gelfoam embolization was infused into all liver segments via transcatheter hepatic arterial placement.
EtOH Ab. 96-proof absolute alcohol, fine-needle, tumor installation was performed with ultrasound guidance, under local anesthesia in cooperative patients with identical peripatient care to TACI. Only tumors ≤2 cm and not abutting Glisson's capsule were considered for this technique.
RFA. All radiofrequency ablations were performed with the interstitial tissue ablation system manufactured by RITA Medical Systems, Inc. (Mountain View, CA). From 2000 on, RITA Model 1500RF Generator and StarBurst™XL were used. The 50-watt generator has a 110–240 V, 50/60 Hertz power supply and visual impedance, time and temperature display. A 3- or 4-array, 15-gauge, 15- or 25-cm length, stainless steel insulated needle with retractable curved electrodes was the energy delivery device used. Each electrode ray has its own thermister to measure temperature changes in the tissue during the ablation. A commercially available ultrasound machine (Aloka, Inc., Wallingford, CT) equipped with linear and sector 7.5–15 MHz probes (for Closed, Open and laparoscopic delivery) was used to guide all ablations. Open or laparoscopic tumor probe placement was necessary only if ultrasound failed to visualize the tumor adequately for percutaneous placement. The size of the tumor determined whether the 3-, 4- or 5-cm ablation company-recommended protocol for target temperature, power setting and time of ablation to a maximum temperature of 110° was used. A 0.5-cm circumferential zone of liver around the tumor ablation was the goal of treatment, requiring 1–2 probe placements for each HCC ablated. All needle tracts were cauterized on withdrawal of RITA probe from the liver, as per the manufacturer's instructions.
Liver explants were taken to the pathology laboratory and sectioned completely in 0.8-cm slides looking for tumor or nodules. When found, the macroscopic aspect of the tumor was compared in size and location with the last half-Fourier RARE MRI pretransplant. Any HCC found further than 0.5 cm from the primary tumor was described as a separate lesion by convention.
All tumor was fixed, microscopically sectioned, and stained with hematoxylin & eosin (H&E). The attending pathologist reviewed all slides. The number of slides containing viable HCC cells was divided by the total number of slides necessary to cover the whole tumor (%), in order to create a semi-quantitative approach to remaining tumor viability.
Patients selected for transplantation were informed prelisting that they would have a back-up candidate to receive ‘their’ donor organ (whether cadaveric or living donor) if HCC was found extrahepatically on exploration for transplant. To date, this has not been necessary. Our standard Cellcept and Neoral or Prograf immunosuppression protocol has been utilized for all HCC patients with two modifications. Steroids are weaned off on all transplanted HCC patients at or before 3 months post transplant (23).
Starting in 2000, rapamycin was substituted for the calcineurin drugs for patients with serum creatinines of ≥2 mg/dL at the time of transplant. No systemic chemotherapy was used pre or post transplant.
Data are expressed as ± standard deviation. The Fisher exact test, the t-test and the Wilcoxon test were used to compare data between the two groups (Table 1). p < 0.05 was statistically significant (Table 1).
Table 1. : Results of HCC treatment groups
|Age (mean)||52.27||57.8||p =0.038*|
|AFP elevated Dx||54%||0%||p =0.012′|
|CTP score (mean)||11±2 points||8±2 points||p<0.01*|
|Time Dx–Tx |
survival at 30months
From December 1997 to July 2001, 182 liver transplants were performed at the Medical College of Virginia. Sixty-one patients (33.51%) received liver grafts from live donors and 121 patients (66.49%) received grafts from cadaveric donors.
The demographic data of the 21 HCC patients under study, revealed a mean age of 54.9 years (44–63 years), 100% male dominance, the same racial mix as the general patient population in the region, and dominance of a viral etiology of the comorbid cirrhosis. In group 1, all HCC patients had cirrhosis, and 10 out of 11 patients had positive viral hepatitis serology. One patient had cirrhosis secondary to α-1-antitrypsin deficiency. In this group of patients the HCC stages were: one T3, eight T2, and two T1at the time of diagnosis (Table 2). The mean CTP score was 11 ± 2 points.
Table 2. : Results of pathologic study of HCC ablated treatment group
|1.||1.1×1.2cm RL||49||T1||neg||1cm 100% necrosis||20m||T0|
|2.||1.5×2cm and 1.4cm RL||51||T2||neg||1.3×1.5cm 100% necrosis; |
new 1.3cm LL
|3.||3.7×2.8cm RL||58||T2||neg||2.5×2.5cm 100% necrosis |
RL; new 1.3cm RL
|4.||5×3.4cm RL||46||T2||neg||2.2×2cm 100% necrosis||9m||T0|
|5.||2×2cm RL||63||T2||neg||1.3×1.2cm 100% necrosis||6m||T0|
|6.||3×3cm RL||60||T2||neg||2cm 95% necrosis RL |
and new 0.8cm RL
|7.||3.8×4cm LL||44||T2||neg||2.5×2cm 70% necrosis||13m||T1|
|8.||2.3×2.3cm RL||45||T2||neg||3.5×2.2cm 40% necrosis||12m||T2|
|9.||1.9×1.5cm RL||47||T1||neg||1.3×1.2cm 95% necrosis||32m||T1|
|10.||6×5cm and 3cm RL||54||T3||neg||5×3cm and 2–1.4cm |
RL 95% necrosis
|11.||3.8×4cm RL||58||T2||neg||3.3×4cm 10% necrosis; |
new 1.2×1.2cm RL
Alpha-fetoprotein was increased in six out of 11 patients (54%) at the time of diagnosis, range 9–12 000 ng/mL, and all patients normalized levels after completing ablation therapy. The safety of the cancer treatment strategy was recorded weekly. The 11 patients in group 1 received 36 ablation procedures for the treatment of 13 liver tumors. The procedures were: TACE (14), TACI (15), PEI (4), and RFA (3). Completing 36 nonresective ablation therapy procedures, one patient developed a surgical complication (1/36, 2.77% morbidity). This was a diaphragmatic perforation found at time of transplant in one patient who underwent RFA for a T1 HCC located in segment four, close to the diaphragm. The patient exhibited mild-to-moderate right hydrothorax before transplant, and the defect was repaired at the time of the transplant. Peri-operative mortality was 0%.
The mean transplant waiting time in this group was 12.91 ± 7.6 months.
Study of explanted livers at pathology showed that 6/11 (54.5%) of ablated tumors had viable HCC, ranging from 5 to 90%, despite all patients' livers lacking enhancement in the last MRI performed before the transplant. Overall, 8 explanted livers (72.72%) had viable cancer cells in this group, 4 livers (36.3%) with new ‘synchronous’ tumors, and 6 livers (54.5%) with incomplete tumor ablation. Five explanted livers (45.5%) showed no viable cancer cells and complete necrosis of the pretransplant ablated tumors.
This group has a post-transplant time mean follow-up of 30 ± 12 months; at this time there has been no patient death, and no HCC recurrence. With the use of multimodality therapy, TNM stages on liver explants at transplant time were: T0N0M0 in three patients, T1N0M0 in six patients, T2N0M0 in one patient and T3N0M0 in one patient. The TNM stage of HCC was reduced in eight patients (72.7%), not changed in two patients (18.2%) and increased in one patient (9.1%).
In this group, two patients received a right lobe liver donation from unrelated donors, and nine (81%) received cadaveric whole livers. Four of 11 patients (36.4%) had an additional synchronous HCC separate from the primary tumor in the explanted liver and were labeled as multifocal HCC.
Clinical group II was represented by 10 HCC patients with stages (three T1, seven T2) with the diagnosis of HCC made at the time of the explanted liver pathologic examination. There was no pretransplant evidence in ultrasound liver studies or tumor markers of HCC. AFP was normal in all patients, mean 3.40 ± 3.43 ng/dL, and all patients had been screened with liver ultrasound before the liver transplant. The mean CTP score in this group was 8 ± 2 points. There were no complications at the time of transplant in this group. There is 0% morbidity and mortality at 30 ± 12 months follow-up and no evidence of HCC recurrence. The statistical analysis is shown in Table 1.
The rising incidence of end-stage liver disease from long-term HCV infection and the improvement of diagnostic techniques have led to a rising incidence of HCC in the last 20 years (4). This climbing incidence of HCC can be expected to grow over the next decades. Knowing the risk potential for cirrhosis-associated hepatocellular carcinoma, the establishment of screening programs, based on liver ultrasonography and the measure of α-fetoprotein levels, in high-risk populations in eastern countries and in patients on waiting lists for liver transplant in western countries has contributed to earlier detection of liver tumors, with improved chances of meaningful treatment (8,15,19). Most of these screening programs have been associated with a lower tumor stage and less severe liver disease at diagnosis. In this study, despite a planned HCC screening program utilizing ultrasound, 50% of the HCC patients (ten T1 and T2 group II incidental HCC patients) were not diagnosed pretransplant. Ultrasound screening did prevent HCCs greater than 5 cm from being transplanted ‘incidentally’ at our center, which has been a reported risk factor for post-transplant recurrence (11). If larger patient numbers prove that pretransplant HCC ablation in fact contributes to prevention of post-transplant cancer recurrence, then the expense vs. benefit of more sensitive screening modalities (i.e. enhanced MRI) in ‘at risk’ cirrhotic populations prior to and during listing for transplantation should be considered. At least, the end-stage cirrhotic recipient fortunate to have a living donor liver option should have a RARE/MRI pretransplant. With HCC lesions identified earlier in an increasing number of patients, along with the development of new techniques of local-disease control for these liver tumors, the approach to the treatment of HCC in cirrhotic patients has changed.
Liver resection for hepatocellular carcinoma has been considered the only curative therapy for small HCCs in eastern countries for the last three decades. Kanematsu et al. reviewed 303 patients with liver resections for HCC performed during a period of 16 years. They found survival at 5 years of 67% and at 10 years of 51%. However, disease-free survival at 5 years was just 27% (24). This high recurrence rate after liver resection can be explained by the multifocal nature of HCC in cirrhotic livers. Belghiti et al. studied the recurrence of HCC after resection of a solitary tumor in 47 patients. Patients were followed with ultrasonography, and at 3 years a 60% incidence of intrahepatic recurrence was found, with 9% of the recurrence around the resection margin (25). Schwartz et al. noted multifocality in 38% of explanted livers after transplantation in patients with <5 cm HCC, and in 82% of the livers of patients with HCCs >5 cm at diagnosis (11). A similar multifocal disease incidence was seen in our treated HCC group (36.3%).
It has long been known that small tumors incidentally found by pathologists rarely recur after cadaveric liver transplantation (24).
Local-disease control techniques used as a treatment for HCC have evolved in the last 20 or 30 years, and their use is center dependent, based on center availability and experience. Radiofrequency ablation (RFA) is a thermal treatment technique designed to produce localized tumor destruction by heating tumor tissue to temperatures that exceed 50 °C. At this temperature, intracellular protein denaturation and the melting of lipid bilayers occur, resulting in direct tumor-cell necrosis. Curley et al. reported on RFA treatment of 149 HCC nodules in 110 patients with a mean follow-up of 19 months: 50.9% of patients were tumor free at the end of the study, and the overall morbidity was 12.7% (26). Although the use of arterial chemoembolization alone has shown no curative survival advantage, as an adjuvant therapy to liver transplant TACE has been well tolerated in our study group, and resulted in a variable degree of tumor necrosis and in some cases complete HCC disappearance (27–29). These data correlate to other published transplant center experiences (5,25). The use of chemoinfusion (TACI) may have resulted in the fibrous encapsulation of HCC tumors noted at histology, contributing to cancer-free survival in our patients at the present time. Similar data have been reported with hepatic chemoinfusion in eastern countries (27–29).
Liver transplantation has become a widely accepted therapy for selected groups of patients with HCC. Increasing numbers of reports are showing the benefit of the use of local-control disease techniques followed by liver transplantation as a therapeutic approach to HCC, with very good outcome and a low incidence of tumor recurrence (10,15–19). Because of the limited supply of cadaveric donor organs, interest has arisen in live-donor liver transplantation (LDLTx) and its use for treatment of HCC. Cheng et al. reported a substantial survival advantage for patients with compensated cirrhosis and small (T1–T2) HCC undergoing LDLTx (30). In our group, two patients received LDLTx and they have had the same outcome at 30 months follow-up as their cadaveric liver transplant peers, but with a significantly shorter transplant waiting-list time. We feel that with more data and long-term follow-up, the benefit of LDLTx in the treatment of HCC will show a survival and cost advantage, because less screening testing is necessary as compared to the longer cadaveric liver waiting-list time. The risk of shorter follow-up of the higher risk T2–T3 HCC patients before LDLTx may be that of undiagnosed extrahepatic metastases appearing post transplant instead of being selected out pretransplant by serial radiologic screening.
In our center the creation of a prospective multimodality HCC diagnostic and treatment plan, carried out by a subgroup of an experienced liver transplant team, and the implementation of a prospective cancer registry for data analysis have resulted in the accurate staging of malignancy and underlying liver disease, yielding optimum patient care. The first and direct proof of this initial statement is realized in the benefit of aggressive HCC screening in the chronic viral hepatitis patients referred and followed by the transplant center. Of our liver transplant patients, 5.49% (10/181 patients) were found incidentally, lower than the incidental volume reported by the International HCC Registry (3). All of our incidental HCC patients were early-stage cancers (T1–T2) with 100% survival and 0% recurrence at 2.5 years actual follow-up. Our findings are similar to the University of Pittsburgh screening experience, and do not support the opinions of the authors of two large HCC transplant studies recommending ‘chemotherapy’ post transplant for these patients, because of poor survival and high HCC recurrence (3,11,31).
All HCC patients in our study group with a mean 30 ± 12 months follow-up are alive with no cancer recurrence. The length of follow-up in this study was 30 ± 12 months. This is notable, as most recurrences of cancer occur within 2 years after transplantation (32,33).
The fact that cancer-free survival in the incidental HCC group was similar (p > 0.05) to a group of T1, T2 and T3 cirrhotic patients treated pretransplant with nonresective ablation techniques is another strong endorsement of the excellent curative results of transplantation when used for selected patients with end-stage liver disease and HCC without systemic, nodal or macrovascular involvement (19). Finally, study size does not allow us to definitively answer ‘the critical question’: does ablation truly down-stage patients so that their HCC-free survival is identical to patients with small incidental HCC? The answer may come in the surrogate question: how many known HCC patients can be maintained with HCC down-staged by ablation until a liver becomes available so that cancer-free survival is identical to that for the ‘small incidental HCC patient’?
We conclude that the use of nonresective ablation techniques is safe and effective in reducing HCC progression in patients with advanced cirrhosis awaiting liver transplantation. This study shows that the combination of multimodality therapy for local-disease control plus liver transplantation in a selected group of cirrhotic patients with HCC results in the same 100% tumor-free survival at 30 ± 12 months as in a group of patients with small incidental (T1–T2) HCCs. Finally, our reported strategy has achieved the goal of optimum treatment of HCC: (i) a strategy of selecting those patients who can be cured by transplantation or resection therapy; (ii) preventing death from liver failure and/or cancer recurrence; (iii) avoiding the use of a scarce organ replacement resource for palliation alone.
Oral presentation in part at American Transplant Congress, Washington, DC, April 28, 2002.