Chemoembolization is the technique of injecting chemotherapy medication into the feeding arteries of a tumor along with particles designed to slow or stop the further arterial supply of oxygen and nutrients to that tumor. It has been performed since the late 1970s.[1],[2] It is one of several techniques used with the goal of treating either primary liver cancer or cancer metastatic to the liver. The most common primary liver cancer is hepatocellular carcinoma (HCC). Common types of metastases to the liver include those from the colon, breast, carcinoid, soft tissue sarcomas, and melanoma. Arterial chemoembolization also can be termed transarterial chemoembolization (TACE).

The liver is divided into lobes and segments; there are several classification schemes for describing these segments, but the most widely used is the Couinaud classification. In this classification, the segments and lobes are defined in part by their third order (third branch) portal venous supply and in part, by their systemic venous drainage.

The arterial liver supply is variable. The most common configuration is that of a single proper hepatic artery derived from the common hepatic artery in turn derived from the celiac artery, with the proper hepatic artery dividing into right and left branches to feed the right and left lobes of the liver, respectively. However, this configuration is probably present in less than 50% of humans; various sources give various percentages. Authors disagree regarding the terminology of the artery that supplies segment 4, which is termed a middle hepatic artery by some and a segment four branch by others.[3] The caudate lobe has particularly variable supply, often with multiple branches from both the right hepatic artery and the left hepatic artery.

The following are the 3 common potential accessory arterial supplies to liver tumors:

• The right phrenic artery (sometimes called the inferior phrenic artery)
• A replaced or accessory left hepatic artery
• The right internal mammary artery (also called the internal thoracic artery)

The following 4 arteries should be kept in mind as common avenues for inadvertent chemoembolization (i.e., embolizing non-target healthy tissues):

• The cystic artery
• The right gastric artery
• The falciform artery
• The duodenal (or supraduodenal) artery

Perhaps the most widely used guidelines for management of cancer in the United States are those created by the National Comprehensive Network (NCCN). The NCCN is a non-profit organization composed of selected hospitals, mostly university hospitals, that choose member physicians from multiple specialties to sit on panels that make national guidelines for cancer treatment.

According to the NCCN, determining whether or not TACE is indicated involves a process of elimination. On the one hand, TACE is not as effective as surgery (either transplant or partial hepatic resection) for curing HCC or metastatic disease and is not in and of itself considered to be a curative treatment. On the other hand, TACE is a more effective treatment than others on the spectrum of treatments that may be considered to be purely palliative. In a minority of situations, TACE either can change a patient's status from "incurable" to "curable" (by downsizing one or more tumors to qualify the patient for surgery) or provide complete permanent tumor necrosis. Thus, on the one hand, a candidate for TACE is someone who has such advanced tumor status and/or overall poor health so as not to be able to withstand surgery or to gain a survival benefit from it as a first line option. This same person, on the other hand, must be healthy enough to withstand the potential side effects and toxicity of TACE and to have good enough odds of gaining at least several months of additional survival at a reasonable quality of life. This process of elimination for discerning which treatment is the best first-line treatment for an individual patient is variable and often subjective, but guidelines do exist to allow for some objectivity.

When considering treatment options for HCC and metastatic disease to the liver, the NCCN relies on several sets of criteria or "scores." These include scores, such as the Child-Pugh-Turcotte score, which is often referred to only as Child-Pugh score, and the United Network for Organ Sharing (UNOS) score, which is commonly referred to instead as the Model for End-stage Liver Disease (MELD) score. The Child-Pugh score is a measure of liver function that depends on a patient's levels of creatinine, INR, total bilirubin, encephalopathy, and ascites. The UNOS score is a measure of liver and renal function. It depends on the creatinine, INR, total bilirubin. A modification of this score called the MELDNa score is currently utilized for organ allocation decision making by UNOS. The MELDNa score is a modification of the MELD score with incorporation of the serum sodium value. It needs to be kept in mind that most of the studies done to assess the safety and appropriate patient selection for chemoembolization is based on MELD score, and not the MELDNa score. 

Additionally, the decision for or against transplant relies on tumor burden. The predominantly used set of criteria to define this are called the Milan criteria, which were established based on a study of 48 patients at the University of Milan.[4]. These criteria state that a patient is a candidate for liver transplant as a preferred first-line treatment for HCC if the liver contains no more than one tumor up to 5 cm in maximum span, or if there are multiple tumors, then there are not more than 3 tumors with each tumor measuring no more than 3 cm in maximum span. Although these are the criteria used by the NCCN, there are alternative tumor burden criteria available, such as from the University of California, San Francisco.

The NCCN usually relies on the American Joint Committee on Cancer staging system to create algorithms for the treatment of various cancers based on cancer staging. However, the AJCC staging system does not take into account liver function, which is a critical component in determining prognosis and potential survival benefit. Multiple other HCC staging systems based on patient cohorts with accumulated survival data have been proposed that estimate the survival of newly diagnosed patients with HCC based on the severity of their tumor burden, liver function, and/or other health factors. HCC has a variable, poorly understood natural history based on its underlying etiology, for example, secondary to hepatitis C versus hepatitis B versus alcohol-induced versus other, and based on its molecular markers. The tumor size and number affect prognosis but may not be as meaningful a predictor of prognosis as the previously mentioned factors. For example, persons in China are relatively more susceptible to developing HCC from hepatitis B, persons in Japan are relatively more susceptible to developing HCC from hepatitis C, and Caucasians are relatively more susceptible to developing HCC from alcohol-induced cirrhosis or steatohepatitis. Consistent with the previous statements, research studies of these different populations that have compared survival in patients with untreated HCC against survival in patients having undergone various treatments for HCC have found moderately different survival estimates. Thus, the prognosis for a person with HCC is best determined by using a study that has evaluated persons with the same set of circumstances. Since staging is designed to correlate with prognosis in order to help decide how aggressive to be with treatment recommendations for a given person, the tumor staging system that is used should ideally be based on patients who correspond most similarly to the patient at hand or at least most closely to the population at that particular hospital.

Available staging systems include but are not limited to the following:

• Okuda, 1985, Japan[5]
• Cancer of the Liver Italian Program, 1998, Italy[6]
• Barcelona Clinic Liver Cancer, 1999, Spain[7]
• Chinese University Prognostic Index, 2002, China[8]
• Japan Integrated Staging, 2003, Japan[9]
• American Joint Committee on Cancer, 2010 United States[10]

A review of the differences, strengths, and weaknesses of these and other staging systems is provided by Kinoshita[11] and in the 2018 NCCN guidelines. There is much overlap between these systems, and the NCCN does not favor the use of one over another. The Barcelona Clinic Liver Cancer staging system (BCLC) is endorsed by the American Association for the Study of Liver Disease (AASLD) andEuropean Association for the Study of the Liver (EASL),  and is likely the most widely used staging system in the United States. It is the focus here and will be discussed in detail. The BCLC itself is a conglomeration of other staging systems, including the Child-Pugh score and the Eastern Cooperative Oncology Group (ECOG) Performance Status (PS).

The Child-Pugh and PST scoring systems and the BCLC treatment algorithm are presented in the graphics below.

In the BCLC, patients are divided into stages 0, A, B, C, and D from least to most severe condition. BCLC stage-B patients are an extremely heterogeneous group, ranging from having four subcentimeter tumors to having one tumor over 5 cm to having a liver to replaced with a tumor. 

A clinical algorithm based on the BCLC and EASL guidelines[12] recommends TACE as the first line treatment if:

• The patient belongs to the Intermediate Stage (Stage B) with preserved liver function and ECOG-PS of zero
• HCC does not fall within resection or transplatation criteria
• In multinodular disease, if the patient has more than 3 nodules, or if any of the tumor size is greater than 3cm when there are greater than or equal to 2 nodules.

The NCCN recommends TACE as first-line treatment if:

• The patient's HCC tumor burden is non-resectable
• The patient's liver function is no worse than Child-Pugh B, and
• The patient's overall function is no worse than PS 2.

The following is a breakdown of patient factors that determine how TACE may be used based on the NCCN guidelines. All patients who are candidates for TACE below are either not a candidate for (ECOG PS greater than 0) or choose against hepatectomy/transplant, or Child-Pugh class A or B.

Patient with HCC

TACE intends to keep the patient a candidate for transplant ("bridging" therapy).

•  No evidence of vascular invasion on imaging and
•  No evidence of extrahepatic spread on imaging and
•  ECOG PS remains 2 or less and
•  Patient meets Milan criteria

TACE intends to downstage the tumor(s) so that the patient qualifies for a transplant.

• As above, but tumor burden is just outside of the Milan criteria

Patient with Metastasis to the Liver

TACE intends to cure.

• All liver and primary lesions are considered potentially curable by thermal/chemical ablation or by resection and
• The patient is willing to be treated as part of a clinical trial

The intent of TACE is to downstage to enable resection.

• All metastatic lesions still qualify for eventual resection and
• Initial chemotherapy has failed and
• The patient is willing to be treated as part of a clinical trial.

The intent of TACE is to palliate.

• At least one lesion does not affect quality for resection regardless of the outcome of TACE and
• Initial chemotherapy has failed and
• The hepatic tumor burden is predominant/is the main cause of symptoms and
• The patient understands that TACE may be of little to no benefit in terms of extending quality of life.

Relative contraindications to TACE include but are not limited to the following:

• Poor expected survival. This can be predicted if the patient has a PST greater than 3, in other words, patient completely disabled, cannot self-care, or confined to a bed or chair. 
• Contraindication to chemotherapy, such as heart or kidney failure and leukopenia
• Unwilling or unable to comply with follow-up guidelines
• Extra-hepatic metastases
• Malignant ascites
• Tumor greater than 5 cm
• Poor baseline liver function (as defined by reference)
• Patients with end-stage cirrhosis (Child-Pugh)
• Encephalopathy
• Jaundice
• Active alcohol consumer
• Hypovascular tumor based on appearance on MRI or CT
• Main portal vein thrombosis associated with a high risk of post-treatment liver failure
• Biliary obstruction or bilirubin greater than 3 mg/dL, unless segmental injections can be performed
• Patient eligible for a potentially more effective therapy

In general, no more than 50% of liver volume should be chemo-embolized at a time. Therefore, if a patient has a tumor occupying more than 50% of the liver volume, then that patient is best treated with 2 TACE procedures.

TACE should not be done in patients with decompensated cirrhosis which is defines as the presence of ascites, encephalopathy, jaundice or variceal bleeding.

TACE usually does not cause significanr damage to the adjacent liver parenchyma univoled with the tumor as HCC receives blood via the hepatic artery ( which is targeted in TACE) and the normal liver parenchyma receives blood mainly from the portal vein. Portal vein thrombosis is not a contraindication to palliative TACE if there is adequate collateral hepatopetal flow to supply the liver parenchyma to be embolized. However, in patients with portal vein thrombosis, it is wise to avoid TACE. Mortality associated withvTACE was reportedly below 1% and most of the deaths were due to liver failure. [13] Thus selecting the right patienrt for this procedure is crucial.

Many physicians treat bile duct obstruction before offering TACE, which could exacerbate bile duct obstruction.

The NCCN does not include TACE in its treatment algorithm for primary cholangiocarcinoma.

TACE can be divided into 2 major categories of chemotherapy delivery:

1. TACE with drug-eluting embolization particles, and
2. TACE in which the active chemotherapy drug and the embolization particles are separate agents (i.e., conventional).

Conventional TACE (cTACE) is a mixture of three agents: lipiodol contrast agent, a chemotherapy drug, and an embolization agent. TACE using drug-eluting beads (TACE-DEB or DEB-TACE) uses resin beads that slowly release chemotherapy. Currently, only two chemotherapy drugs are FDA-approved for use in this manner. Doxorubicin is used to treat hypervascular primary hepatic tumors and metastases. Irinotecan is used to treat metastases from colorectal cancer. While TACE with drug-eluting beads has similar efficacy and fewer side effcts compared to conventional TACE, the use of bland embolization is controversial.