Metastatic disease of the liver is more common than primary liver neoplasms. The liver is the most common site of metastatic deposits after the lymph nodes.1 The top three primary sites with liver metastases are breast, lung and colorectal.2 Transarterial chemoembolisation (TACE) has been successfully used to treat unresectable hepatocellular carcinoma (HCC) and similar principles are being applied in the treatment of metastatic liver disease.
The liver has a dual blood supply with the portal vein contributing about 75% of the total and the remaining 25% being supplied by the hepatic artery.3 The preferential hepatic arterial supply to the tumours of liver4 allows selective arterial embolisation without injury to the hepatic parenchyma, which is supplied by the portal vein.
Embolisation with iodised oil and gel foam results in lower peak plasma concentration of the drug and increased retention of the drug within the tumour as compared with hepatic arterial administration of drug alone.5 This could result in prolonged action of the drug on the tumour and the reduced systemic levels, which means less toxicity. Ischemia induced by embolisation also augments the action of the chemotherapeutic agents.
Patient selection, inclusion, exclusions and contraindications
Patients that have metastatic liver disease with isolated liver disease and no extra hepatic disease or limited/stable extra hepatic disease are candidates for TACE. Patients must have either unresectable liver lesions, lesions not responding to systemic chemotherapy or with toxicity to systemic chemotherapy. With neuroendocrine tumours metastasising to the liver, chemoembolisation can be performed in the presence of extra hepatic disease to reduce the hormonal symptoms or mass effect from the liver lesions. If the tumour burden exceeds 70% of the liver volume, chemoembolisation is avoided as this can lead to hepatic failure. Patients must be in the Eastern Cooperative Group (ECOG) grade 0 or 1. The presence of systemic diseases like lung or skeletal metastases are a contraindication. Complete thrombosis of the portal vein is a contraindication to TACE; however in presence of partial thrombosis with adequate hepatopetal collaterals, a case-based approach can be considered. Biliary dilation in the setting of TACE predisposes the patient to biliary strictures and abscesses and decompression is indicated before TACE. General contraindications to angiography include renal failure (serum creatinine >2mg/dl), hepatic failure (total serum bilirubin >3mg/dl, serum albumin <2g/dl), coagulation abnormalities and contrast allergy.
All patients receive an MRI of the liver with a multiphasic study to document disease burden, hepatic arterial anatomy and also to evaluate any possible extra hepatic disease. Patency of the portal vein is evaluated in the pre-procedure study. If the portal vein is partially thrombosed, the presence of collaterals is noted. For follow up TACE, a non-contrast MRI of the upper abdomen including a T2, T1 and diffusion weighted study is performed. The primary and the metastatic liver lesions are evaluated with biopsy and histopathology confirmation is obtained. Baseline renal and liver function tests and carcinoembryonic antigen (CEA), in the case of colorectal cancer, are performed close to the procedure. Adequate counselling of the patient and family is done regarding the nature of the procedure and the expected results.
Fasting for at least six hours before the procedure and admission on the day of the procedure is followed. Premedication with antiemetics and corticosteroids is provided along with adequate saline infusion. Analgesic coverage is provided in the procedure based on patient tolerance, generally with pethidine/morphine.
Femoral arterial access is secured with a 5 Fr sheath and Seldinger technique. The 5 Fr pigtail catheter is used to obtain an aortic angiogram to delineate the arterial anatomy and variants. The celiac trunk is cannulated with a 5 Fr cobra or sidewinder catheter. Selective angiograms of the superior mesenteric artery are also obtained to evaluate for possibility of replaced right hepatic artery (Figure 1). Delayed image acquisition with the superior mesenteric artery angiograms is done to establish the patency of the portal vein. An exchange length angled tip guide wire is used throughout the procedure to facilitate catheter exchange. An appropriate sized microcatheter like the Renegade Hi FLO (inner diameter 0.27”) is used to obtain selective angiograms of the right and left hepatic systems (Figure 2) and also to delineate the gastroduodenal and cystic arteries. The large inner diameter of the microcatheter system also facilitates in delivery of drug without catheter blockage.
Figure 1: Patient with carcinoma pancreas post surgery (clips). Aortogram (A) and selective right hepatic artery angiogram (B) showing replaced right hepatic artery arising from the superior mesenteric artery.
Figure 2: Selective right hepatic angiography with sidewinder catheter in the celiac trunk and microcatheter in the right hepatic artery showing two hypervascular lesions with the top one showing good lipiodol uptake.
Non-hepatic arterial supply to the lesions is to be recognised, as this is likely to result in a suboptimal response to hepatic arterial chemoembolisation. Specific signs such as subcapsular location of the lesion and hypertrophied arteries surrounding the lesion are to be sought for because they predict the likelihood of extra hepatic arterial supply.
Following the angiograms, the lobe with the largest lesion or the lobe with largest burden of metastatic disease is chosen for embolisation. Nearly a third to half of our patients have an ultrafast cone beam computed tomography (CT) image acquisition during the study as it helps accurately identify the vessels feeding the tumour and guides the catheter tip position.
Chemotherapeutic drugs used are mitomycin C (up to 10mg/m2), cisplatin (up to 35mg/m2) and gemcitabine (up to 1000mg/m2). The choice of drugs is based on discussions in the tumour board about the use of mitomycin C alone, mitomycin C with gemcitabine, mitomycin C with irinotecan or mitomycin C with cisplatin and gemcitabine. Lipiodol, up to 10ml, followed by starch microspheres 200–450mg are used for embolisation. The end points of embolisation are determined when stasis of flow in the target vessels is noted. In patients with slow flow states caution must be exercised to avoid over embolisation and non-target embolisation. The arterial puncture site is closed with a vascular closure device or manual compression.
Patients are observed in the observation room for six hours following the procedure with monitoring of vital observations and puncture site. Then 1000ml of saline is infused over the period. In case the of no adverse events patients are discharged the same evening. All patients receive proton pump inhibitors for four weeks following the procedure.
All patients have a non-contrast CT scan three hours after the procedure to evaluate for distribution of lipiodol (Figure 3). Repeat imaging and TACE is performed after an interval of at least four weeks. Tumour marker levels like that of CEA in case of colorectal metastasis are measured to document reduction.
Figure 3: Axial CT scan after chemoembolisation of the right lobe of liver for metastatic liver disease showing good lipiodol uptake in multiple lesions of the right lobe of liver.
The incidences of complications related to the arterial puncture, such as puncture site haematoma and pseudoaneurysm formation, are common to all procedures and are few in number. There is a small risk of thrombosis and dissection of the hepatic artery especially when it is very tortuous. This can be avoided by meticulous care to the technique. The fact that close to half of the patients have hepatic arterial variations of non-target embolisation is an important consideration. Non-target embolisation of the vessels supplying the stomach and intestine could lead to ulcerations and perforation. Embolisation of the pancreaticoduodenal vessels can result in pancreatitis. Accidental embolisation of the cystic artery leads to chemical cholecystitis, which is often self-resolving without requiring cholecystectomy. Transient raise in the liver enzymes is common following TACE. The incidence of liver failure is low if patients are carefully selected based on patency of portal vein and Child-Pugh status.
Post-embolisation syndrome manifests following TACE as right upper quadrant pain, fever, nausea and vomiting. There may be a rise in the leukocyte count; however, routine antibiotic administration is not justified. Symptomatic and conservative management is necessary in most patients. Liver abscesses are more commonly seen in patients with biliary enteric anastomosis.
Colorectal tumours: Tellez et al. conducted a phase II trial for metastatic colorectal carcinoma to the liver. They included 30 patients who were resistant to other therapies and underwent up to three sessions of TACE with bovine collagen material, cisplatin, doxorubicin and mitomycin C. Treatments were repeated at six to eight week intervals. The response was measured by reduction in the density of the lesion or reduction in the size. They reported a response in 63% of the cases and also reduction in CEA in 95% of the cases. Median survival was 8.6 months from the start of TACE and 29 months after diagnosis of hepatic metastasis.6 Albert et al. studied the role of TACE in colorectal liver metastasis with cisplatin, doxorubicin, mitomycin C, ethiodol and polyvinyl alcohol with about one to four cycles of therapy. They reported a partial response in 2%, stable disease in 41% and progression in 57%. Survival from the time of diagnosis of primary colon cancer was 33 months, 27 months from diagnosis of liver metastasis and nine months after initiation of TACE. They also reported better survival if TACE was initiated after first- or second-line systemic chemotherapy rather than after third- to fifth-line therapies. In their study, the presence of extra hepatic metastatic disease did not affect survival adversely.7 In our department, we carried out a prospective study to evaluate local tumour control and survival after TACE with different drug combinations in palliative treatment of liver metastases in colorectal cancer.
We included 463 patients and they underwent 2441 sessions of TACE. The chemotherapy regimens included mitomycin C alone (243 patients), mitomycin C with gemcitabine (153 patients) and mitomycin C with irinotecan (67 patients). Lipiodol and starch microspheres were used as embolising agents. RECIST criteria were used to evaluate tumour response. Partial response was noted in 14.7%, stable disease in 48.2% and progressive disease in 37.1% of patients. The one- and two-year survival after TACE was 62% and 28% respectively. Median survival from diagnosis of liver metastases was 38 months and that from the start of TACE was 14 months. There was no survival advantage of one chemotherapeutic protocol over another.8 Another study from our institute, which aimed to determine the prognostic factors leading to treatment success in colorectal cancer with liver metastases treated with TACE has demonstrated that the indication for treatment and initial tumour response were significant factors affecting the patient’s survival. TN stage, presence of extra hepatic metastases, number of lesions, tumour location within the liver and chemotherapy protocol were not significant factors determining the patient’s survival.9
Neuroendocrine tumours: Gupta et al. reported their results with hepatic arterial embolisation and chemoembolisation for metastatic neuroendocrine tumours in 69 patients. They concluded that patients with carcinoid tumours had better outcomes than those with islet cell tumours. There was no improvement with the addition of intra-arterial chemotherapy to hepatic artery embolisation in patients with carcinoid tumours, but had benefits in patients with islet cell tumours. An intact primary tumour, extensive liver disease and bone metastases were associated with reduced survival in patients with islet cell tumours.10
Ho et al. studied the outcome of TACE and embolisation of hepatic metastatic lesions from neuroendocrine tumours (carcinoid (31 patients) and islet cell tumours (15 patients)). A total of 46 patients had 93 embolisation sessions. Mean overall survival was 1273 ±185 days. There was no statistically significant difference in survival between the carcinoid and islet cell tumour subgroups. The progression free survival times for both these subgroups were also similar. There was significant difference in the survival of patients with and without extra hepatic disease (770 ±112 days versus 1571 ±291 days). There was no statistically significant difference in subgroups with and without resection of the primary tumour. In 80% of the patients who had hormonally active tumours (20 out of 25 patients) there was relief in the symptoms after one cycle of chemotherapy.11
Breast cancer: Cho et al. retrospectively reviewed 10 patients with breast cancer who were treated with TACE for liver unresectable metastases.12 These patients had failed first and second line systemic therapy. The drugs used for TACE were adriamycin (n=6), cisplatin/gemcitabine (n=2) cisplatin (n=1) and oxaliplatin (n=1). The median number of TACE cycles was four. They reported an increase in the median survival for patients who had responded to treatment when compared to those who did not respond (24 versus seven months). They concluded that TACE was a feasible option and needed further investigation.
Our department has reported the largest study available to date where we studied the efficacy of TACE with different drug combinations in patients with breast cancer and liver metastases with respect to local tumour control and survival.13 208 patients with unresectable hepatic metastases of breast cancer underwent TACE at four-week intervals. For all drug protocols together, their reported local tumour control was partial response in 13%, stable disease in 50.5% and progressive disease in 36.5%.
The one-, two- and three-year survival rates after TACE were 69, 40 and 33%, respectively. The mitomycin C alone protocol showed median and mean survival times of 13.3 and 24 months, and the same was seen with gemcitabine, except it was only 11 and 22.3 months. With the combination of mitomycin C and gemcitabine, survival times were 24.8 and 35.5 months leading them to conclude that better results were obtained with combination chemotherapy.
Martin et al. studied the role of TACE with drug-eluting beads in metastatic breast cancer.14 40 patients with metastatic breast cancer spreading to the liver underwent TACE with hepatic arterial drug-eluting beads loaded with doxorubicin. Hepatic progression-free survival was a median of 26 months and overall survival was a median of 47 months. They concluded that treatment of hepatic metastases from metastatic breast cancer using doxorubicin-eluting beads was an effective local therapy with very high response rates. They recommended that in comparison to chemotherapy alone, consideration of hepatic directed therapy be warranted in patients with liver dominant metastatic disease.
Cholangiocarcinoma: Burger et al. studied 17 patients with unresectable cholangiocarcinoma with TACE and reported median survival time of 23 months for TACE. After TACE, two patients who were previously unresectable became resectable.15 Kim and colleagues analysed 49 patients with inoperable intrahepatic cholangiocarcinoma treated between 1997 and 2007 with either TACE or transcatheter arterial infusion (TACI). Following the procedure, 55% of the patients showed radiographic response. They also inferred that tumour vascularity was the only independent factor associated with radiographic response. Large tumour size, hypovascular tumours and Child-Pugh class B were poor prognostic factors for patient survival.16 Park et al. compared the survival benefits of TACE with supportive therapy; 72 patients underwent TACE and 83 patients underwent supportive therapy.
The survival period was longer in the TACE group being 12.2 months versus the supportive therapy group with a survival period of 3.3 months.17 A study from our institute with the aim of evaluating effectiveness of TACE with four different chemotherapeutic protocols (mitomycin C only, gemcitabine only, mitomycin C with gemcitabine and combination of gemcitabine, mitomycin C and cisplatin) with respect to local tumour control and survival was done for 115 patients who underwent a total of 819 chemoembolisations at four-week intervals (mean 7.1 sessions per patient). Local tumour partial response was observed in 8.7%, stable disease in 57.4% and progressive disease in 33.9% of patients. Initial tumour response, high tumour vascularity and Child-Pugh class A were statistically significant factors for patient survival. There was no statistically significant difference between different chemotherapeutic protocols.18
Ovarian cancer: Liver parenchymal metastases from ovarian cancer are less common than peritoneal metastases. Hepatic parenchymal involvement is seen in 2–12.5% of patients with ovarian cancer. Vogl et al. studied the role of repetitive TACE as third-line therapy of ovarian cancer metastasis to the liver.19 Patients in this study had cytoreductive therapy for the primary ovarian tumour before TACE. They reported a partial response in 16.9%, stable disease in 58.5% and progressive disease in 24.6% of patients. However they noted no difference in median and mean survival times between subgroups using mitomycin C alone, mitomycin C with gemcitabine or mitomycin C with gemcitabine and cisplatin.
Transarterial chemoembolisation of hepatic metastases is useful method of treating patients with systemic chemotherapy resistance; it helps prolong the progression-free survival and reduces the symptoms of neuroendocrine tumours. It is associated with few complications if proper patient selection, evaluation and technique are adhered to.
Currently a variety of different interventional oncology techniques are developed for the treatment of malignant liver tumours. Among them, transarterial approaches like transarterial chemoembolisation (TACE), transarterial chemoperfusion (TACP), transarterial embolisation (TAE) are the most commonly performed procedures. Radioembolisation using different radioactive drugs is currently further developed. Thermoablative techniques like laser ablation (LITT), microwave ablation (MWA), radiofrequency ablation (RFA) and radiotherapy are another important part in interventional oncology for the treatment of malignant liver tumours. In future, major efforts will be directed towards individualised, more controlled therapies in liver cancer as well as molecular and genetic therapy modalities.
1 Disibio G, French S. Metastatic patterns of cancers: Results from a large autopsy study. 2008, Arch Pathol Lab Med 2008;132:931–39.
2 Ananthakrishnan A et al. Epidemiology of primary and secondary liver cancers. Semin Intervent Radiol 2006;23:47–63.
3 Vollmar B, Menger M. The hepatic microcirculation: Mechanistic contributions and therapeutic targets in liver injury and repair. Physiological Rev 2009;1269–339.
4 Breedis C, Young G. The blood supply of neoplasms in the liver. Am J Pathol 1954;969–85.
5 Raoul J et al. Chemoembolisation of hepatocellular carcinomas: A study of the biodistribution of doxorubicin. Cancer 1992; 585–90.
6 Tellez C et al. Phase II trial of chemoembolisation for the treatment of metastatic colorectal carcinoma to the liver and review of the literature. Cancer 1998;1250–9.
7 Albert M et al. Chemoembolisation of colorectal liver metastases wIth cisplatin, doxorubicin, mitomycin C, ethiodol and polyvinyl alcohol. Cancer 2011;117:343–52.
8 Vogl T et al. Repeated transarterial chemoembolisation in the treatment of liver metastases of colorectal cancer: Prospective study. Radiology 2009;281–9.
9 Gruber-Rouh T et al. Transarterial chemoembolisation of unresectable systemic chemotherapy refractory liver metastases from colorectal cancer: Long term results over a 10 year period. Int J Cancer 2014;134:1225–31.
10 Gupta S et al. Hepatic arterial embolisation and chemoembolisation for the treatment of patients with metastatic neuroendocrine tumors. Cancer 2005;104:1590–602.
11 Ho AS et al. Long term outcomes after chemoembolisation and embolisation of hepatic metastatic lesions from neuroendocrine tumours. Am J Roentgen 2007;188:1201–7.
12 Cho SW et al. Transcatheter arterial chemoembolisation Is a feasible palliative locoregional therapy for breast cancer liver metastases. Int J Surg Oncol 2010;2010:251621.
13 Vogl T et al. Transarterial chemoembolisation (TACE) with mitomycin C and gemcitabine for liver metastases in breast cancer. Eur Radiol 2010;20:173–80.
14 Martin R et al. Optimal outcomes for liver-dominant metastatic breast cancer with transarterial chemoembolisation with drug-eluting beads loaded with doxorubicin. Breast Cancer Res Treat 2012;132:753–63.
15 Burger I et al. Transcatheter arterial chemoembolisation in unresectbale cholangiocarcinoma: initial experience in a single institution. J Vasc Intervent Radiol 2005;16:353–61.
16 Kim JH. Transcatheter arterial chemoembolisation or chemoinfusion for unresectable intrahepatic cholangiocarcinoma. Cancer 2008;113:1614–22.
17 Park S-Y et al. Transarterial chemoembolisation versus supportive therapy in the palliative treatment of unresectbale intrahepatic cholangiocarcinoma. Clin Radiol 2011;66:322–8.
18 Vogl T et al. Transarterial chemoembolisation in the treatment of patients with unresectable cholangiocarcinoma: results and prognostic factors governing treatment success. Int J Cancer 2012;131:733–40.
19 Vogl T el al. Initial experience with repetitive transarterail chemoembolisation (TACE) as a third line of treatment of ovarian cancer metastasis to the liver. Indications, outcomes and role in patient’s management. Gynecologic Oncol 2011;124:225–9.