INTRODUCTION — Hepatocellular carcinoma (HCC) is an aggressive tumor that frequently occurs in the setting of cirrhosis. The two factors that are most important in determining a patient's prognosis and potential treatment options are the tumor mass and the patient's hepatic reserve.
Treatment options are divided into potentially curative surgical therapies (ie, resection and orthotopic liver transplantation) and nonsurgical therapies, some of which may provide long-term disease control (ie, radiofrequency ablation [RFA]; microwave ablation [MWA]; percutaneous ethanol injection [PEI]; cryoablation; irreversible electroporation; transarterial embolization, which includes bland embolization, chemoembolization, and radioembolization; radiation therapy; and systemic therapy). In general, for most patients with HCC limited to the liver who are not candidates for surgical resection or liver transplantation, locoregional liver-directed therapies are preferable to systemic therapy.
This topic review will cover transarterial embolization and RT. Other ablative nonsurgical therapies (RFA, MWA, PEI, cryoablation, microwave and laser coagulation, irreversible electroporation) are discussed elsewhere, as are resection, liver transplantation, and systemic therapy for HCC, as well as an overview of treatments for HCC. (See "Localized hepatocellular carcinoma: Liver-directed therapies for nonsurgical candidates who are eligible for local ablation" and "Surgical management of potentially resectable hepatocellular carcinoma" and "Liver transplantation for hepatocellular carcinoma" and "Systemic treatment for advanced hepatocellular carcinoma" and "Overview of treatment approaches for hepatocellular carcinoma".)
TREATMENT ALGORITHMS AND GENERAL APPROACH TO THE PATIENT WITH LOCALIZED DISEASE — An algorithmic approach to the treatment of HCC is shown in the figure (algorithm 1). The suggested approach is useful for conceptualizing the various treatment options that are available for individual patients, but it may not be applicable in all settings. Alternative algorithms are available from other groups, such as the one that is used by the Barcelona Clinic Liver Cancer group; their most recent updated algorithm is available online . However, attempts to generate algorithmic approaches to the treatment of HCC are difficult since new treatments and indications for various treatments are evolving rapidly. Furthermore, therapeutic approaches tend to vary based upon underlying liver function, available expertise, and variability in the criteria for hepatic resection and orthotopic liver transplantation. These issues are discussed in detail elsewhere. (See "Overview of treatment approaches for hepatocellular carcinoma".)
IMPORTANCE OF MULTIDISCIPLINARY CARE — The wide variety of treatments for HCC are offered by different specialties: surgery, radiation oncology, medical oncology, and interventional radiology. Multidisciplinary evaluation and planning typically result in more thoroughly vetted recommendations and are less likely to result in a recommendation for a procedure for which a single provider has expertise but that may not represent the optimal therapy for an individual patient.
The majority of patients with HCC have underlying liver disease. Patients who undergo any form of therapy for HCC are at high risk for progression to liver failure because of their underlying liver disease, and proper monitoring, assessment, and treatment of the underlying liver disease may have a major impact on long-term survival. Multidisciplinary care has been shown to improve survival in patients with HCC [2,3]. The comprehensive care of patients with cirrhosis includes antiviral therapy for hepatitis B and C virus, immunization against hepatitis A and B virus (if indicated), regular surveillance for HCC with abdominal imaging, and endoscopic screening and surveillance for varices. (See "Epidemiology and risk factors for hepatocellular carcinoma" and "Cirrhosis in adults: Overview of complications, general management, and prognosis" and 'Issues related to HBV reactivation' below.)
ASSESSING RESPONSE TO LOCOREGIONAL THERAPIES — The antitumor effect of many nonsurgical locoregional treatment modalities for HCC is not accurately reflected by conventional bidimensional tumor measurements performed on radiographic studies. Accurate assessment of response following nonsurgical locoregional treatment requires evaluation of residual tumor enhancement after therapy as based on the American College of Radiology's Liver Reporting and Data System (LI-RADS) treatment response algorithm after locoregional therapies for HCC (algorithm 2) [4,5]. The most important feature of the treatment response algorithm is persistent lesion contrast enhancement and washout. For some patients, the effectiveness of therapy may also be monitored by serial assay of tumor markers such as alpha fetoprotein (AFP). These same characteristics are used for early detection of residual/recurrent tumor and new areas of tumor involvement. (See "Assessment of tumor response in patients receiving systemic and nonsurgical locoregional treatment of hepatocellular cancer".)
Assessment of treatment response after nonsurgical local therapies should be performed with multiphase (including arterial, portal venous, and delayed phase imaging), contrast-enhanced, cross-sectional imaging (computed tomography [CT] with iodinated contrast or magnetic resonance imaging [MRI] with gadolinium-based agents). MRI is the preferred imaging modality with a sensitivity and specificity for lesion detection of 88 and 94 percent, respectively . As tumor enhancement characteristics post-therapy are essential in assessment for tumor viability , contrast should be administered unless there are contraindications. Ordering the examinations specifically for HCC post-treatment surveillance imaging will ensure an appropriate imaging protocol will be used. Patients with iodinated contrast allergies should be routed to MRI and those with gadolinium allergies to CT. Preexisting kidney impairment is not considered a contraindication for contrast-enhanced MRI .
We recommend dynamic (contrast-enhanced) cross-sectional imaging with either CT or MRI one to three months after nonsurgical locoregional therapy for HCC. Using the LI-RADS treatment response algorithm, tumors are categorized as viable, nonviable, or equivocal [4,8]. The absence of contrast uptake within the tumor reflects nonviable tumor, while the persistence of contrast uptake indicates persistent disease, referred to as viable. If a tumor has enhancement atypical for treatment, it is deemed equivocal. Recurrence of tumor in the treated area (or elsewhere) is signaled by the reappearance of vascular enhancement or the development /growth of enhancing soft tissue along the periphery of the treated lesion. (See "Assessment of tumor response in patients receiving systemic and nonsurgical locoregional treatment of hepatocellular cancer", section on 'Response after locoregional therapy'.)
We typically reimage every three months for two years. After two years with no evidence of disease recurrence, we revert back to standard HCC surveillance with cross-sectional imaging and AFP assay every six months. At many institutions, cross-sectional imaging every six months is continued beyond the two-year cutoff, given the continued elevated risk for HCC in these patients. However, after two years with no evidence of disease recurrence, other institutions revert back to standard HCC surveillance with ultrasound and AFP assay every six months, especially for low-risk patients who are not transplant candidates. The use of ultrasound in this situation is controversial, and many believe that the risk remains sufficiently high that periodic cross-sectional imaging should be continued beyond two years. We agree with this approach. (See "Surveillance for hepatocellular carcinoma in adults", section on 'Summary and recommendations'.)
PATIENTS WITHOUT PORTAL VEIN THROMBUS
Embolization — Hepatic arterial embolization is an appropriate option for patients with a large unresectable or multifocal HCC who have relatively preserved liver function (ie, Child-Turcotte-Pugh class A or B cirrhosis (table 1)) and no extrahepatic tumor spread, vascular invasion, or tumor thrombus involving the main portal vein or one of its lobar branches. It is also appropriate for "bridging" therapy in a patient awaiting liver transplantation and for selected patients prior to resection of a large HCC, particularly involving the right lobe, in conjunction with portal vein embolization (PVE).
The majority of the blood supply to a HCC is derived from the hepatic artery rather than the portal vein. This has led to the development of techniques designed to eliminate the tumor's blood supply by particle embolization and/or to directly infuse cytotoxic chemotherapy into the branch of the hepatic artery that feeds the tumor.
Hepatic arterial embolization, using bland embolization, chemoembolization, or radioembolization, is a reasonable option for patients with an unresectable HCC that is either too large or multifocal for percutaneous ablation techniques such as radiofrequency ablation (RFA) or microwave ablation (MWA); with relatively preserved liver function (ie, Child-Turcotte-Pugh class A or B cirrhosis (table 1)); and with no extrahepatic tumor spread, vascular invasion, or tumor thrombus involving the main portal vein or one of its lobar branches. It is also appropriate for "bridging" therapy in patients awaiting liver transplantation and for selected patients prior to resection of a large HCC, particularly involving the right lobe, in conjunction with PVE. Patient selection is critical to the success and safety of transarterial embolization. Benefit from any of these treatments is highly dependent upon tumor-related factors and the severity of the preexisting liver dysfunction.
Is there an optimal embolization approach? — There is marked variability in the types of procedures that have been used for transarterial embolization: bland particle embolization; transarterial chemotherapy, alone or using drug-eluting beads (DEB) and with or without lipiodol; transarterial chemoembolization (TACE), with and without other local therapies (ie, radiation therapy [RT] or tumor ablation); and radioembolization [9-11]. Few randomized trials have directly compared different techniques for hepatic artery embolization, and there is little consensus as to the best approach. For patients in whom hepatic arterial embolization is indicated who do not have portal vein thrombus, we suggest TACE rather than bland embolization or radioembolization. TACE is the only method of transarterial treatment that has been shown in randomized trials to provide a survival advantage compared with supportive care alone for patients with a large unresectable HCC that is not amenable to liver transplantation or local ablation, and it is the most commonly used technique.
While additional data from randomized trials are ideally needed to resolve the question of whether TACE plus local ablation is better than either procedure alone, the available evidence supports combined therapy, and at many institutions, TACE plus local ablation (RFA, MWA) is preferred over TACE alone for tumors >2 cm. We agree with consensus-based guidelines from the National Comprehensive Cancer Network (NCCN), which support the use of RFA or MWA plus TACE for intermediate-sized HCC (ie, 3 to 5 cm) . (See 'Unresectable or multifocal HCC' below and 'Guidelines from expert groups' below.)
Given the potential for treatment-related toxicity, we reserve the combination of TACE plus RT for patients with portal vein thrombus but preserved unilateral portal blood flow. (See 'TACE plus radiation therapy' below.)
Meta-analyses comparing some of these techniques have had mixed results [13-16].
A network analysis comparing the effectiveness of different transarterial therapy techniques in 55 randomized trials with 5763 patients who had preserved liver function and unresectable HCC came to the following conclusions :
●All of the embolization strategies achieved a significant survival gain over control treatment (hazard ratio [HR] range 0.42 to 0.76, very low to moderate quality of evidence).
●TACE, TACE with DEBs (DEB-TACE), radioembolization, and TACE plus systemic antitumor therapy did not confer any survival benefit over bland embolization alone (moderate quality of evidence, except low quality of evidence for radioembolization).
●There was moderate-quality evidence that TACE combined with RT or local ablation achieved the best survival. The estimated median survival was 13.9 months in the control group, 18.1 months with TACE, 20.6 months with DEB-TACE, 20.8 months with bland embolization, 30.1 months with TACE plus RT, and 33.3 months with TACE plus local ablation. Objective response rates were higher when TACE was combined with RT or ablation compared with TACE alone.
●With regards to treatment-related serious adverse effects, radioembolization was the safest treatment (odds ratio [OR] for serious adverse events compared with the control arm 6.35, 95% CI 1.11-5.56). The least safe treatments were TACE plus RT (OR for serious adverse events 53.1) and TACE plus systemic therapy (OR 68.5).
DEB-TACE has become the most commonly used approach to hepatic arterial embolization at many institutions, including those of the authors; bland embolization is not widely accepted. At many institutions, TACE plus local ablation is a preferred approach over TACE alone for tumors >2 cm, although the quality of the data is low. (See 'Drug-eluting beads' below and 'Radioembolization' below.)
Guidelines from expert groups — Guidelines from expert groups differ. TACE, rather than bland embolization alone, is recommended in guidelines from an expert consensus group of the Americas Hepato-Pancreato-Biliary Association (AHPBA) . On the other hand, guidelines from both the NCCN  and the American Association for the Study of Liver Diseases (AASLD)  recommend chemoembolization, bland embolization, or radioembolization in this setting, with neither group preferentially recommending one form of locoregional therapy over another. (See 'Radioembolization' below.)
TACE and bland particle embolization — TACE is an appropriate option for patients with a large unresectable or multifocal HCC without main or lobar branch portal vein thrombus that is not amenable to local ablation. TACE is also commonly used as a bridging maneuver in patients awaiting liver transplantation. For most patients who are candidates for resection, preoperative TACE is not indicated. However, TACE may be a complementary procedure prior to PVE if a major liver resection is planned, particularly for a right-sided lesion. (See "Liver transplantation for hepatocellular carcinoma", section on 'Chemoembolization' and "Surgical management of potentially resectable hepatocellular carcinoma", section on 'Portal vein embolization' and "Management of potentially resectable hepatocellular carcinoma: Prognosis, role of neoadjuvant and adjuvant therapy, and posttreatment surveillance", section on 'Neoadjuvant therapy'.)
Bland particle embolization, which relies solely on induction of tumor ischemia by disruption of the blood supply to the tumor, has been utilized successfully for the treatment of both initially unresectable and recurrent HCC [19-22]. However, the majority of the published experience is with TACE, in which transarterial embolization is combined with injection of chemotherapeutic agents, with or without lipiodol (which is sometimes called transarterial oily chemoembolization), into the hepatic artery. Although the evidence from comparative trials is weak [21,23,24] and meta-analyses comparing both procedures have come to differing conclusions [13-16], conventional TACE is the only method of transarterial treatment that has been shown in randomized trials to provide a survival advantage compared with supportive care alone for patients with a large unresectable HCC that is not amenable to liver transplantation or local ablation [23,25], and it is the most commonly used technique for hepatic arterial embolization. (See 'Is there an optimal embolization approach?' above.)
Patient selection and contraindications — The best candidates for TACE are patients with unresectable HCC without vascular invasion or extrahepatic spread and with compensated liver function (ie, no worse than Child-Turcotte-Pugh class A or B cirrhosis (table 1)). Baseline liver function seems to be the most important factor predicting survival in patients with unresectable HCC treated with TACE [26-28].
While all patients with portal vein thrombus have been excluded from TACE in the past, there is now some evidence that patients who have thrombus not involving the main portal vein or its major branches and have preserved unilateral portal blood flow may tolerate the procedure [26,29,30]. However, complete portal vein occlusion is associated with a worse outcome with TACE alone than partial or segmental occlusion. Management of these patients is discussed below. (See 'Patients with portal vein invasion or thrombus' below.)
Absolute contraindications to TACE include absent portal vein flow and decompensated cirrhosis (Child-Turcotte-Pugh class C, or Child-Turcotte-Pugh class B score >8 including jaundice, clinically overt hepatic encephalopathy, refractory ascites, and/or hepatorenal syndrome) [31,32].
Relative contraindications include a variety of other factors, including but not limited to:
●Serum bilirubin >2 mg/dL
●Lactate dehydrogenase >425 units/L
●Aspartate aminotransferase >100 units/L
●Tumor burden involving >50 percent of the liver
●Untreated esophageal varices at high risk of bleeding
●Prior transjugular intrahepatic portosystemic shunting (TIPS)
Although several small retrospective case series have demonstrated a good safety profile for TACE among patients with prior TIPS, we suggest that patients be carefully selected and excluded if they have marginal hepatic function. Furthermore, if TACE is undertaken in this setting, selective tumor embolization should be done, if at all possible, in order to minimize hepatic ischemia [33-36].
New clinical scoring systems have been developed to improve patient selection for TACE (eg, the Barcelona Clinic Liver Cancer B subclassification system, the Hepatoma Arterial Embolization Prognostic [HAP] score, the Selection for TransArterial chemoembolization TrEatment score [STATE]) [37-39].
As an example, one of these, the HAP score, is derived by assigning one point each to the following factors: albumin <36 g/dL, alpha-fetoprotein (AFP) >400 ng/mL, bilirubin >17 micromol/L (1 mg/dL), and maximum tumor diameter >7 cm . Median survival for individuals undergoing TACE or transarterial embolization with 0, 1, 2, or >2 points prior to the embolization procedure (corresponding to HAP score A, B, C, and D groups) was 27.6, 18.5, 9.0, and 3.6 months, respectively. The performance of this score in patients with larger tumors who are treated with combined TACE plus sorafenib has not been addressed.
Unresectable or multifocal HCC — TACE is most often used for patients with relatively preserved liver function who have unresectable or multifocal HCCs that are not amenable to other local treatments. Approximately 35 to 40 percent of patients have an objective antitumor response (≥25 percent reduction in tumor size) [23,40-43]. However, many tumors do not decrease in size even after successful treatment, forcing the use of surrogate markers of response (lack of contrast enhancement on computed tomography [CT] or magnetic resonance imaging [MRI], lipiodol deposition in targeted tumors, decline in AFP) to determine the presence of tumor necrosis and the likelihood that treatment has been beneficial [44-46]. When criteria are used that take into account the area of intratumoral necrosis to estimate the reduction in tumor burden, the rate of objective response is as high as 60 percent [13,47]. When using criteria such as these, sustained response durations of six months or longer are an independent prognostic factor for prolonged overall survival . (See 'Assessing response to locoregional therapies' above.)
Despite the widespread acceptance of TACE and the suggestion of prolonged survival in multiple phase II studies, three of five published randomized trials failed to show a survival advantage for TACE compared with a variety of other treatments or conservative management in patients with advanced unresectable HCC [42,43,49]. On the other hand, two relatively small randomized trials have shown a survival advantage for TACE compared with symptomatic treatment alone in selected patients with unresectable HCC and preserved liver function [23,25]. By contrast, no trial has found a survival advantage for bland embolization alone. (See 'Is there an optimal embolization approach?' above.)
Four meta-analyses have addressed the survival benefit of TACE relative to control treatments or bland embolization, with disparate results, likely due to selection bias [13-16]. As an example, in one of the randomized trials comparing TACE with symptomatic treatment alone, TACE significantly prolonged one-year survival (82 versus 63 percent), but survival in the control group was markedly higher than that reported in most treatment studies for advanced HCC .
Prior to resection — A number of uncontrolled series and at least one controlled trial suggest that TACE used prior to an attempt at resection is associated with increased mortality. Thus, TACE is generally not indicated prior to potentially curative resection. (See "Management of potentially resectable hepatocellular carcinoma: Prognosis, role of neoadjuvant and adjuvant therapy, and posttreatment surveillance", section on 'Neoadjuvant therapy'.)
One possible exception is in patients who need major resection (eg, right hepatectomy). Preoperative PVE is a valuable adjunct to major liver resection, particularly for right-sided tumors. TACE and radioembolization may be complementary procedures prior to PVE in such patients as they eliminate the arterial blood supply to the tumor and also embolize potential arterioportal shunts that attenuate the effects of PVE in cirrhotic livers. This topic is discussed in detail elsewhere. (See "Surgical management of potentially resectable hepatocellular carcinoma", section on 'Portal vein embolization'.)
Prior to liver transplantation — Despite the granting of Model for End-Stage Liver Disease (MELD) exception points for liver transplant candidates with HCC whose tumors are within the Milan criteria, potential recipients must wait for six months before getting their exception points, and most wait at least an additional six months to a year for a donor organ (unless they are candidates for living donor transplantation). (See "Liver transplantation for hepatocellular carcinoma", section on 'Requirements for listing and management while on the wait list' and "Liver transplantation for hepatocellular carcinoma", section on 'Living donor transplantation'.)
Several small series have demonstrated the feasibility of TACE prior to orthotopic liver transplantation in patients who are estimated to have a long wait time to transplantation. Despite its uncertain benefit, it remains a common method in many transplant centers to limit the amount of viable tumor at the time of transplantation to help bridge the often prolonged waiting period while a patient is listed and awaiting a donor organ and to improve selection for transplantation based upon tumor biology. In addition, TACE may be used to downstage larger tumors so that they are within Milan criteria and can be listed for transplant. This topic is discussed in detail elsewhere. (See "Liver transplantation for hepatocellular carcinoma", section on 'Chemoembolization' and "Liver transplantation for hepatocellular carcinoma", section on 'Downstaging through neoadjuvant locoregional therapy'.)
Complications — The most common adverse effect of TACE, which occurs in 60 to 80 percent of patients, is postembolization syndrome. This consists of varying degrees of right upper quadrant pain, nausea, a moderate degree of ileus, fatigue, fever, and transient elevation of aspartate aminotransferase, alanine aminotransferase, and bilirubin values. Symptoms are usually self-limited, lasting three to four days; full recovery is typical within 7 to 10 days. This may be less common among patients treated with DEBs rather than other methods of embolization, such as gelatin sponge (Gelfoam).
Some suggest that tumor necrosis is the main cause [42,50], while more recent reports attribute postembolization syndrome to ischemic damage to normal liver parenchyma [51,52]. Prospective studies are needed to definitively answer the question of whether postembolization syndrome is a side effect to be avoided or an index of the efficacy of the procedure.
Although most of the chemotherapy is retained in the liver, there is some systemic exposure, and patients are at risk for nausea, vomiting, and bone marrow depression. The use of DEBs reduces systemic exposure to chemotherapy. (See 'Drug-eluting beads' below.)
Other, potentially more serious, complications of TACE include the following:
●Treatment-induced ischemic damage to the non-tumor-bearing liver is thought responsible for precipitating or exacerbating liver failure. In the meta-analysis cited above, liver failure rates averaged 7.5 percent, but the wide range (0 to 49 percent) likely reflects the heterogeneous patient population .
The incidence of hepatic decompensation following TACE is closely related to pretreatment hepatic function [40,53,54]. In a prospective study of 197 sessions of TACE performed in 59 patients with HCC, acute hepatic decompensation developed after 39 sessions (20 percent) and was irreversible in six cases (3 percent) . Patients with irreversible changes were significantly more likely to have higher pre-TACE bilirubin levels, more prolonged prothrombin time, and more advanced cirrhosis.
More often, hepatic decompensation is reversible, but it may be irreversible. In one study of 251 consecutive patients with HCC and synthetic hepatic dysfunction who underwent 443 TACE procedures, reversible hepatotoxicity developed in 78 patients (31 percent), while irreversible hepatotoxicity developed in 37 patients (15 percent) . Risk factors for irreversible hepatotoxicity were serum bilirubin ≥4 mg/dL, prolonged prothrombin time, serum albumin <2 g/L, serum creatinine >2 mg/dL, large ascites, encephalopathy, or a MELD score ≥20. Notably, these are all patient populations for whom TACE would be considered contraindicated. (See "Model for End-stage Liver Disease (MELD)" and 'Patient selection and contraindications' above.)
●Other less common ischemic complications include hepatic abscess (2 percent ), acute cholecystitis, and injury to the biliary tract. Bile duct injury (subcapsular biloma, focal stricture of the hepatic or common bile duct, diffuse dilation of the intrahepatic bile ducts) is reported in 0.5 to 2 percent of cases [14,56].
●Gastroduodenal ulceration is a recognized complication of TACE (3 to 5 percent) [14,57]. Among the possible explanations are regurgitation of embolizing particles into the right or left gastric artery, the presence of anatomic variants (eg, the right gastric artery too distal to the proper hepatic artery, or an accessory left gastric artery arising from the left hepatic artery), or stress ulcers.
●Renal dysfunction (2 percent) is also reported.
●Pulmonary and cerebral lipiodol embolization are rare but potentially fatal [58-60].
●An interstitial pneumonia is also described in which radiographic studies are not suggestive of lipiodol retention in the lung . Histologic evaluation (and the fact that most cases occurred after a second injection) raised the possibility of an immunoallergic reaction, with a possible contribution from radiation pneumonitis. However, hypereosinophilia was absent, and glucocorticoid therapy seemed to be of little value. Of note, 12 of the 15 patients reported in this series died, mostly of acute respiratory failure.
●Treatment-related mortality rates from TACE are generally less than 1 percent, but higher rates (approximately 2 to 3 percent) are reported, mostly in patients with very large tumors in whom tumor lysis syndrome develops after TACE [14,62].
Issues related to HBV reactivation — TACE is a risk factor for reactivation of hepatitis B virus (HBV) infection, and antiviral prophylaxis is recommended for patients who are hepatitis B surface antigen positive. At least some data suggest better outcomes among patients undergoing TACE for HBV associated HCC who are treated concomitantly with nucleoside analogs (lamivudine, adefovir, or entecavir) [63,64]. Issues related to HBV screening and prophylaxis are described separately. (See "Hepatitis B virus reactivation associated with immunosuppressive therapy".)
Technique for TACE — The methodology for TACE varies; the term has been used to refer to the injection of a chemotherapeutic agent into the hepatic artery with or without lipiodol and with or without procoagulant material. Lipiodol is an oily contrast agent that is thought to promote intratumoral chemotherapy retention; however, it probably does not contribute to arterial occlusion [14,65]. While it is thought that chemolipiodolization and embolization have synergistic antitumor activities (the former by inducing high intratumoral concentrations of cytotoxic drugs and the latter by cutting off the blood supply to the tumor), this has never been proven. Worldwide, the most commonly used chemotherapy agents for TACE to treat HCC are doxorubicin, epirubicin, or cisplatin; few comparator trials are available .
In theory, embolization should enhance the effect of chemotherapy by causing metabolically active cell membrane pumps to fail, thereby overcoming drug resistance . However, whether simultaneous or sequential occlusion of the hepatic artery until there is stagnation of blood flow to the tumor results in greater antitumor efficacy than chemotherapy alone has not been proven. Neither the optimal chemotherapeutic agent nor the best embolization method (ie, gelatin sponge [Gelfoam], degradable starch microspheres, polyvinyl alcohol) has been established [14,68].
Drug-eluting beads — Where available, we suggest TACE with drug-eluting beads (DEB-TACE) over conventional TACE. The available data suggest that DEB-TACE is associated with a better side effect profile, although long-term disease-control is likely similar.
A newer approach to TACE, which is being used at most centers, uses DEBs that slowly release chemotherapy, theoretically reducing treatment-related toxicity. In addition, the embolic material (the drug-coated beads) remains within the arteries, in contrast to the lipiodol used in most TACE procedures, which may cross the hepatic sinusoids into the portal venules, causing ischemia and a "dual embolic hit" to the liver parenchyma . DEB-TACE has become the most commonly used approach to hepatic arterial embolization at many institutions, including those of the authors. (See 'Is there an optimal embolization approach?' above.)
Results from several small prospective randomized trials and a meta-analysis of seven randomized trials of TACE versus DEB-TACE suggest similar rates of tumor control as with conventional TACE but lower rates of postprocedure pain and serious hepatobiliary toxicity, although follow-up is short in most series [70-78]. On the other hand, a 2017 network analysis comparing various methods of hepatic arterial embolization concluded that TACE and DEB-TACE were associated with a similar risk of serious adverse effects . There is no evidence that DEB-TACE improves survival over conventional TACE .
In our view, the best data on comparative toxicity come from the largest randomized trial, the PRECISION V trial, in which conventional TACE using doxorubicin (50 to 75 mg/m2) was directly compared with DEB-TACE (150 mg of doxorubicin per procedure) in 212 patients with Child-Turcotte-Pugh class A/B cirrhosis and unresectable HCC . The DEB-TACE group had lower rates of treatment-emergent adverse events in the hepatobiliary system (16 versus 25 percent) . The mean maximum postchemoembolization alanine transaminase increase with DEB-TACE was 50 percent less than in the conventional TACE group (p <0.001), and the mean maximum aspartate transaminase increase was 41 percent lower. Furthermore, despite a higher mean total dose of doxorubicin in the DEB-TACE group (295 versus 233 mg), there was a small but statistically significant difference in mean change from baseline in the left ventricular ejection fraction of four percentage points that favored the DEB-TACE group. The incidence of postembolization syndrome was similar between both groups (25 versus 26 percent for DEB-TACE and conventional TACE). On the other hand, treatment-emergent gastrointestinal adverse events occurred more often in patients treated with DEB-TACE (61 versus 45 percent).
Doxorubicin-eluting beads are approved in Europe and Canada (DC Bead, Biocompatibles International, Inc). In the United States, the LC Bead (AngioDynamics, Inc) is commercially available in a variety of sizes and is approved as a medical device. At least some data from a nonrandomized study suggest enhanced efficacy and fewer complications with the smallest particle size (100 to 300 microns) as compared with larger beads .
Irinotecan-eluting beads are beginning to be studied, but only limited data are available, particularly regarding toxicity [81-83].
Procedure — The TACE procedure starts with catheterization of the hepatic artery followed by identification of the tumor's feeding vessel. The selectivity of the catheter placement determines the amount of liver embolized. For patients with solitary tumors, the catheter can be placed into a second-order (selective) or third-order (superselective) branch off the right or left hepatic artery . In cases where there are multiple tumors in one lobe, lobar chemoembolization may be needed; however, this is associated with a higher risk of hepatic ischemia and postembolization syndrome. Whenever possible, selective or superselective TACE is preferred. If there is disease involving both lobes, bilateral TACE must be performed sequentially (we generally wait four weeks between procedures). Whole-liver chemoembolization must be avoided because of the potential for serious liver injury and subsequent hepatic decompensation.
Once the feeding vessel is selected, the chemotherapy is then infused. The choice of chemotherapeutic agent is not standardized, and a variety of agents have been used [85-91]. Some centers use single-agent doxorubicin 60 mg plus lipiodol (20 mL), emulsified with water-soluble contrast (10 mL). Others recommend a mixture of cisplatin (50 to 100 mg), doxorubicin (20 to 50 mg), and mitomycin (10 mg), mixed in 10 mL of water-soluble contrast and then emulsified in an equivalent volume of lipiodol [62,84]. Doxorubicin DEBs are used preferentially at many institutions. (See 'Drug-eluting beads' above.)
In the event that powdered doxorubicin is not available, equivalent doses of the premixed solution should be used, and an attempt should be made to minimize the overall volume. Regardless of the specific chemotherapy "cocktail" used, the resulting emulsion must be constantly shaken during the infusion, which is then followed by 0.1 to 0.2 mL of 150 to 250 micron polyvinyl alcohol dry particles or gelatin sponge pledgets .
We administer broad-spectrum intravenous antibiotics and antiemetics (a serotonin receptor antagonist, dexamethasone, and diphenhydramine) just prior to the procedure. Following TACE, patients are hospitalized until oral intake is adequate and pain is well controlled. In our experience, most patients are discharged within 48 hours of the procedure.
The most important components of postprocedure care include aggressive hydration (3 L per 24 hours), continued antibiotics (parenteral for 24 hours), prophylaxis against nausea and vomiting (typically prochlorperazine, with a serotonin receptor antagonist and dexamethasone only as needed), adequate narcotic analgesia (parenteral or oral as needed), and monitoring of electrolytes and liver function tests. Whether prophylactic antibiotics are needed after discharge is controversial. Some recommend five days of oral broad-spectrum coverage , while others conclude that prophylactic antibiotics are not routinely necessary but are recommended in patients with biliary reconstruction and impaired liver function . We typically administer five days of an oral broad-spectrum antibiotic (eg, ciprofloxacin) after discharge only if the patient developed noninfectious fevers within 24 hours after the procedure. (See 'Complications' above.)
As noted above, to assess the efficacy of TACE, we typically perform dynamic multiphasic cross-sectional imaging and assay AFP (if initially elevated) initially at four weeks, two to three months later, and then at three-month intervals for at least two years. (See 'Assessing response to locoregional therapies' above.)
TACE plus local treatments
●TACE plus local ablation – Consistent with consensus-based guidelines from the NCCN , we suggest the use of RFA or MWA plus TACE rather than TACE or RFA/MWA alone for intermediate-sized HCC (ie, 3 to 5 cm). Given the potential for treatment-related toxicity, we reserve the combination of TACE plus RT for patients with portal vein thrombus but preserved unilateral portal blood flow. (See 'TACE plus radiation therapy' below.)
Combining local ablation treatment (eg, RFA, MWA, or RT) with TACE can theoretically overcome the limitations of either TACE or ablation when used alone. The chance of achieving complete (>90 percent) necrosis is presumably higher with combined treatment, and some uncontrolled series note better survival with combined treatment over TACE alone [94,95]. However, the quality of the evidence to support combined therapy is generally low . While additional data from randomized trials are ideally needed to resolve the question of whether TACE plus local ablation is better than either procedure alone, the available evidence supports combined therapy, and at many institutions, TACE plus local ablation (RFA, MWA, RT) is preferred over TACE or ablation alone for tumors >2 cm.
The following data are available to support the benefit and toxicities of combined therapy with TACE plus local ablation:
•Several randomized trials and at least three meta-analyses have concluded that combined TACE plus RFA improves survival compared with RFA or TACE alone [16,96-98]. In the most recent network analysis, which included 55 trials that both directly and indirectly compared treatments for unresectable HCC, TACE plus RFA was associated with the highest median survival (33 compared with 18.1 months with TACE alone) and had a more favorable rate of serious adverse events compared with TACE alone . However, in all cases, the quality of the evidence was judged to be low. (See "Localized hepatocellular carcinoma: Liver-directed therapies for nonsurgical candidates who are eligible for local ablation", section on 'RFA plus TACE'.)
•Fewer data are available on the combination of TACE plus MWA. Retrospective single-institution analyses suggest that results with TACE plus MWA are at least as good as those with TACE plus RFA. There are no randomized trials. The superiority of combined treatment over TACE alone was suggested in a retrospective report comparing outcomes in 258 patients with a large solitary or multinodular HCC who underwent TACE plus MWA (n = 92) or TACE alone (n = 166) at a single institution over a four-year period [99-101]. At a median follow-up of 21 months, the one-, two-, and three-year survival rates with TACE plus MWA were 86, 60, and 33 percent; they were 59, 40, and 11 percent for TACE alone. The corresponding recurrence rates were 48, 78, and 95 percent for TACE plus MWA and 75, 96, and 98 percent for TACE alone.
●TACE plus RT – Given the potential for treatment-related toxicity, we reserve the combination of TACE plus RT for patients with portal vein thrombus but preserved unilateral portal blood flow. (See 'TACE plus radiation therapy' below.)
RT can be safely delivered to the liver, especially using advanced techniques such as three-dimensional conformal radiation therapy (3D-CRT) and stereotactic body radiation therapy (SBRT) approaches. (See 'Stereotactic body radiation therapy' below.)
•A survival benefit for combined therapy has been suggested in several analyses [16,94,102,103]. As an example, a meta-analysis of 16 trials of TACE with or without external beam RT concluded that combined therapy was significantly better than TACE alone for treatment of HCC . However, a subsequent Cochrane analysis of nine of these trials concluded that the quality of the evidence to support this conclusion was low to very low, and that additional high-quality trials were needed to assess further the role of RT for unresectable HCC .
•Likewise, a Cochrane review of eight randomized trials of TACE plus 3D-CRT compared with TACE alone also concluded that combined therapy "may have reduced" all-cause mortality at three years of follow-up, but that the quality of the evidence was low to very low .
Data from the University of Michigan show that SBRT can be an effective treatment for patients with portal vein thrombosis, with tumor regression in 75 percent of patients and stability in another 17 percent in one small series . Historical data using older RT techniques had objective response rates to RT alone for patients with portal vein thrombosis in the range of 25 to 58 percent [105-112]. Data for TACE plus RT compare favorably with that for sorafenib alone in this population. (See 'Patients with portal vein invasion or thrombus' below.)
Retreatment — Chemoembolization procedures using DEBs are typically repeated for each lesion, one month apart. Otherwise, we do not generally repeat the TACE procedure unless there is clear evidence of progressive tumor growth or residual viable tumor in the treated areas. The decision to retreat with TACE for clear-cut disease progression must be individualized and based upon an assessment of tumor response and the hepatic reserve. In addition, the interventional radiologist's assessment of degree of arterial stasis following embolization can help determine the need for future procedures.
Overall, less than 2 percent of patients achieve a complete response from a single round of treatment with TACE, but this is variable, with higher response rates in smaller tumors. During follow-up, residual tumor nests recover their blood supply and begin to regrow. This has prompted many centers to repeat the TACE procedure at regular intervals in an attempt to maximize benefit . However, in our view, the benefit of multiple repeated cycles of TACE must be weighed against the adverse effects of repeated procedures, particularly in those with cirrhosis and poor hepatic reserve, and we limit TACE to the minimum number of procedures necessary to control the tumor.
Each course of TACE causes some degree of ischemic hepatic damage, which if repeated, has the potential to lead to hepatic decompensation. Excess deaths from deterioration of liver function may counterbalance any prolongation of survival that results from enhanced tumor control . (See 'Complications' above.)
At least two prognostic scoring systems have been developed, based upon post-treatment liver function and radiologic response, to identify patients whose prognosis is too poor after a single TACE session to justify additional TACE treatments [39,114,115]. However, neither of these models has been independently validated.
Repeated courses of therapy may not even be possible. TACE causes hepatic artery damage, the likelihood of which is higher in patients with impaired liver function [116,117]. This has the potential to limit the ability to carry out repeated procedures. Even if repeated TACE procedures are feasible, hepatic artery interruption from repeated TACE or arterial dissection can lead to the development of extrahepatic collateralization, which may create an alternative blood supply to the tumor and contribute to treatment failure .
The definition of "failure of local therapy" for HCC is evolving. The rapidly changing availability of potentially effective and survival-prolonging systemic therapies in HCC has enabled clinicians to offer alternatives to patients who have locally progressive intrahepatic disease, even if they may be eligible for additional liver-directed therapy, such as TACE. Specifically, the acceptance of the toxicity of a risky local procedure in a patient with borderline indications may be tempered by having viable alternative systemic options that did not previously exist. (See "Overview of treatment approaches for hepatocellular carcinoma", section on 'Treatment at progression' and "Systemic treatment for advanced hepatocellular carcinoma".)
Radioembolization — Radioembolization using intra-arterial injection of yttrium-90 (90Y)-labeled glass or resin microspheres induces extensive tumor necrosis with an acceptable safety profile. However, there are no studies demonstrating an impact on survival and no consensus as to the optimal use of this therapy, particularly when and if it should be chosen over TACE for treatment of unresectable HCC. One clinical scenario in which radioembolization may be preferred over TACE is in the setting of an HCC complicated by malignant main or lobar branch portal vein thrombus. (See 'Patients with portal vein invasion or thrombus' below.)
Where available, and for appropriately selected patients, newer techniques for superselective radioembolization such as segmental radioembolization (also called radiation segmentectomy) may provide high rates of local control with less radiation-induced liver disease (RILD) [119-124]. (See 'Complications' below.)
Selection for radioembolization requires evaluation by a multidisciplinary team. Benefits over other forms of nonsurgical locoregional therapy include relatively low toxicity, the potential to treat patients with significant tumor burden (often in a single setting rather than multiple sessions, as for classic TACE ), and relatively limited side effects. However, high cost and certain anatomical constraints (eg, pass through of the radioactive material to the lung in some patients with shunting) limit the utility of this treatment. (See 'Indications and efficacy' below.)
Contraindications — An absolute contraindication to 90Y microsphere treatment is a pretreatment technetium-99m macroaggregated albumin scan that demonstrates the potential for ≥30 Gy radiation to be shunted to the lungs or inadvertent flow to the gastrointestinal tract that cannot be corrected with catheter techniques. Another contraindication is prior RT involving the liver. Otherwise, contraindications to radioembolization are similar to those for TACE and include:
●Clinically overt hepatic encephalopathy
●Child-Turcotte-Pugh class C cirrhosis (table 1)
Relative contraindications include bilirubin >2 mg/dL and ascites. Unlike TACE, main portal vein thrombus or obstruction is not a contraindication to radioembolization. There is little actual embolization in the sense of causing arterial ischemia after radioembolization based upon the size of the beads, so the portal blood supply is less important from the standpoint of toxicity. Radioembolization also appears to be safe in patients with prior TIPS . (See 'Patients with portal vein invasion or thrombus' below.)
Indications and efficacy — Few high-quality trials have compared radioembolization with other embolization strategies, and there is a lack of consensus as to the utility of radioembolization, as illustrated by the differing recommendations from expert groups:
●2018 updated clinical practice guidelines from the AASLD do not recommend one form of embolization over any other; updated guidelines from the European Association for the Study of the Liver (EASL) state that the subgroup of patients benefitting from radioembolization needs to be defined [18,126].
●Guidelines from the NCCN include chemoembolization, bland embolization, and radioembolization as acceptable treatment modalities for patients with liver-isolated HCC who are not candidates for curative therapy .
●An expert consensus group of the AHPBA concluded that 90Y radioembolization could be considered for treatment of HCC in the following scenarios :
•Downstaging/bridging to transplantation or resection
•Portal vein thrombus
Prior to resection — Preoperative PVE is a valuable adjunct to major liver resection, particularly for right-sided tumors. As with TACE, transarterial radioembolization may be a complementary procedure prior to PVE in such patients as it eliminates the arterial blood supply to the tumor and also embolizes potential arterioportal shunts that attenuate the effects of PVE in cirrhotic livers . This topic is discussed in detail elsewhere. (See "Surgical management of potentially resectable hepatocellular carcinoma", section on 'Portal vein embolization'.)
Unresectable primary HCC — Radioembolization using 90Y-tagged glass (TheraSphere) or resin (SIR-Spheres) microspheres is safe and effective in patients with unresectable HCC, including those with portal vein thrombus. (See 'Radioembolization' below.)
Both products are commercially available in North America, while the resin product is available worldwide. In the United States, 90Y glass microspheres were approved for treatment of unresectable HCC in March 2021, while 90Y resin microspheres are approved for treatment of colorectal cancer liver metastases.
Radioembolization is safe and effective for primary treatment of unresectable HCC, even in those with portal vein thrombus [128-137]. The long-term efficacy of 90Y radioembolization is best demonstrated by the following two studies:
●In a comprehensive analysis of 291 patients treated with TheraSpheres , the World Health Organization response rate (objective bidimensional measurement of perpendicular tumor size) was 42 percent; by EASL criteria (measuring the area of enhancement ), 57 percent of patients responded (23 percent complete). (See 'Assessing response to locoregional therapies' above.)
At a median follow-up of 31 months, the median time to progression was 7.9 months. For patients with Child-Turcotte-Pugh class A and B cirrhosis (table 1) without portal vein thrombus, it was 15.5 and 13 months, respectively, while for those with Child-Turcotte-Pugh class A and B cirrhosis and portal vein thrombus, it was 5.6 and 5.9 months, respectively. Median survival time was longer for Child-Turcotte-Pugh class A rather than class B cirrhosis (17.2 versus 7.7 months). Median survival was poor for those patients with Child-Turcotte-Pugh class B cirrhosis and portal vein thrombus (2.6 months). The main clinical toxicities were nausea, abdominal pain, fatigue, and transient hyperbilirubinemia; safety results were not reported separately for those patients with and without portal vein thrombus.
●The multicenter Local radioEmbolization using Glass microspheres for the Assessment of tumor Control with Y-90 (LEGACY) study included 162 consecutively treated patients undergoing radioembolization with glass microspheres for a solitary HCC ≤8 cm, no worse than Child-Turcotte-Pugh class A cirrhosis, and an Eastern Cooperative Oncology Group performance status of 0 to 1 . Radioembolization served as bridging (neoadjuvant) therapy prior to transplantation in 34 (21 percent) and surgery in 11 (6.8 percent). At a median follow-up of 29.9 months, the confirmed radiographic objective response rate was 72 percent, and the median duration of response was 11.8 months. At 24 months, there were no patients with local progression. Three-year overall survival was 87 percent overall, and 93 percent for those patients undergoing neoadjuvant therapy.
Grade 3 events occurred in 31 patients (19 percent), and no patient experienced RILD despite the use of higher radiation doses than previously administered using radioembolization (median absorbed dose to the treated liver volume was 410 Gy) . Hepatobiliary disorders included ascites in three patients, gallbladder obstruction in one, and portal vein thrombus in one. Serious adverse events occurred in approximately 10 percent of patients, and 5.6 percent were at least possibly related to radioembolization.
Largely based on this study, Y90 glass microspheres were approved by the FDA for treatment of unresectable HCC in March 2021.
Whether results with radioembolization are better than those that can be achieved with TACE is unclear; the available data are conflicting:
●One small pilot randomized trial in 28 patients suggests similar safety and disease control rates with a single session of 90Y radioembolization as with multiple sessions of TACE .
●On the other hand, benefit for 90Y radioembolization over TACE was suggested in a subsequent small randomized phase II study in which 45 patients (out of a potentially applicable pool of 179 potentially eligible patients who met the enrollment criteria) were randomly assigned to a single dose of conventional TACE or 90Y radioembolization . The study was discontinued prematurely because of slow accrual before reaching its goal of 124 patients. At a median follow-up of 17.2 months, the estimated median time to tumor progression with radioembolization was significantly longer (estimated >26 versus 6.8 months with TACE), but the response to therapy was similar, as was the median survival time censored to liver transplantation (18.6 versus 17.7 months). The small size of this study and the lower-than-expected survival complicate interpretation of these data.
●A year 2017 network analysis of various hepatic arterial embolization methods concluded that transarterial radioembolization did not confer any survival benefit over bland transarterial embolization alone  but that it was the safest option compared with TACE alone or in conjunction with RT or systemic chemotherapy; however, the quality of the evidence was low. (See 'Is there an optimal embolization approach?' above.)
Similarly, a year 2020 Cochrane analysis concluded that the evidence showing effects of radioembolization versus chemoembolization in unresectable HCC was highly insufficient ; radioembolization did not seem to differ from TACE in terms of serious adverse events or quality of life, but the certainty of evidence was very low.
None of these analyses included patients treated with superselective radioembolization, which may provide similar or better outcomes as nonselective radioembolization or selective chemoembolization but with less RILD [119-124], and few studies used personalized dosimetry, which improves outcomes as compared with standard dosimetry .
Patients being considered for liver transplantation — Although there are less data than with TACE, radioembolization is safe and limits disease progression, which may allow patients more time to wait for a donor organ [142-146]. In addition, patients who present with tumors that are beyond the usual transplantation criteria and who do not have malignant portal vein thrombus or extrahepatic disease involvement may be candidates for radioembolization to downstage the disease to within transplant criteria [147,148]. (See "Liver transplantation for hepatocellular carcinoma", section on 'Radioembolization' and "Liver transplantation for hepatocellular carcinoma", section on 'Downstaging through neoadjuvant locoregional therapy'.)
Complications — Complications occurring after radioembolization include the following [149-152]:
●A mild postembolization syndrome with fatigue, constitutional symptoms, and abdominal pain (incidence 20 to 70 percent). This is generally less severe than after TACE.
●Hepatic dysfunction due to radioactivity to the surrounding liver parenchyma may manifest as liver failure (attributed to hepatic sinusoidal injury ) or RILD (defined as jaundice and ascites appearing one to two months after radioembolization in the absence of tumor progression or bile duct occlusion [153-156]).
There are only limited data on the hepatic parenchymal response to radioembolization. Among the changes that are described in patients receiving lobar infusions are transient hyperbilirubinemia, ipsilateral hepatic lobar volume decrease with contralateral lobar hypertrophy (a phenomenon that has been termed "radiation lobectomy"), induction of liver fibrosis, and portal hypertension [157,158].
The frequency and risk factors for development of these adverse effects, and their influence on liver insufficiency have not been well studied. These issues were addressed in a retrospective series of 260 patients with liver tumors (27 with HCC, 67 with colorectal cancer, 25 with neuroendocrine tumors, and 47 with other tumors) treated with radioembolization, 75 using a standard protocol and 185 using a modified protocol that included ursodeoxycholic acid (300 mg twice daily for two months starting on the day of radioembolization) plus methylprednisolone (8 mg daily for one month and 4 mg daily for a second month after radioembolization) . RILD appeared only in patients with cirrhosis or in noncirrhotic patients who were exposed to systemic chemotherapy within two months before radioembolization. The incidence of all-stage RILD was reduced from 23 to 5 percent with use of the modified protocol, while the incidence of severe RILD was reduced from 13 to 2 percent. Disease-control rates were identical in patients treated with the standard and the modified protocol.
Prior exposure of the liver to RT may lead to increased liver toxicity after radioembolization; radioembolization should be restricted to those patients who have had limited or no hepatic exposure to RT [152,159].
Although the data are limited, superselective or segmental radioembolization (radiation segmentectomy) may provide similar efficacy as nonselective radioembolization for appropriately selected patients, with a better side effect profile [119-123].
●Hepatic fibrosis and/or portal hypertension are potential complications [158,160]. These may become more obvious as patients survive longer, and/or the published incidence may underestimate the frequency of these complications due to relatively short follow-up. However, clinically significant manifestations, such as thrombocytopenia or variceal bleeding, are rarely seen following radioembolization.
●Radiation pneumonitis is very rare (incidence <1 percent) unless the lung shunt fraction is high, in which case the chance of delivering a high pulmonary dose increases.
●Gastric or duodenal injury manifests as severe pain during the procedure. The incidence is <5 percent if proper percutaneous techniques are used. Pretreatment angiography is essential to identify and coil embolize variant vessels that may supply the gastrointestinal tract.
●Vascular injury, which is most often seen in patients receiving chemotherapy.
●Lymphopenia ; the majority of patients have a >25 percent decrease in lymphocyte count post-treatment.
Response assessment — In general, patients treated with radioembolization have a relatively delayed tumor response; the median time to develop evidence of necrosis (decreased enhancement) and tumor shrinkage is approximately 30 and 120 days, respectively. For this reason, we typically perform the initial radiographic assessment of disease response at approximately three months post-treatment rather than one month later, as is typically done for other forms of nonsurgical local treatment. Postradiation changes are often difficult to distinguish from residual disease complicating post-treatment interpretation. The American College of Radiology's Liver Reporting and Data System (LI-RADS) reporting system is being used to standardize reporting of these post-treatment findings (algorithm 2). (See 'Assessing response to locoregional therapies' above.)
Retreatment — There are few data on retreatment with 90Y radioembolization and no information on the number of safe treatments. At least one study suggests that the risk of potentially fatal radioembolization-induced liver disease is increased after a second treatment .
Is there a role for systemic therapy? — There has been a resurgence of interest and enthusiasm for systemic therapy of HCC with the emergence of data showing that the molecularly targeted agents sorafenib and regorafenib improve survival compared with best supportive care alone; a survival benefit has also been shown for combined atezolizumab plus bevacizumab compared with sorafenib in the first-line setting, and in the second-line setting for nivolumab, an immune checkpoint inhibitor, lenvatinib, has also demonstrated noninferiority to first-line sorafenib. (See "Systemic treatment for advanced hepatocellular carcinoma".)
For patients with liver-isolated HCC who are eligible for liver-directed nonsurgical therapies, two relevant questions are whether the addition of systemic therapy improves results compared with locoregional therapy alone, and whether initial systemic therapy provides better outcomes than can be achieved with initial liver-directed therapy. Taken together, the available evidence from limited randomized trials does not support a clear benefit for the addition of sorafenib to TACE and also suggests that embolization outperforms initial sorafenib in terms of efficacy and adverse event profile. This position is now being reevaluated in light of the benefits of immune checkpoint inhibitors, and several ongoing trials are exploring combined therapy.
A separate but related question is when should patients with liver-isolated disease be considered to have a "failure of local therapy," for which systemic treatment should be initiated. The rapidly changing availability of potentially effective and survival-prolonging systemic therapies in this disease has enabled clinicians to offer alternatives to patients with locally progressive intrahepatic disease. Specifically, the acceptance of the toxicity of a risky local procedure in a patient with borderline indications may be tempered by having viable alternative systemic options that did not previously exist. (See "Systemic treatment for advanced hepatocellular carcinoma".)
This subject is discussed in detail elsewhere. (See "Overview of treatment approaches for hepatocellular carcinoma", section on 'Treatment at progression'.)
TACE plus sorafenib — For patients who are eligible for TACE, we suggest not adding sorafenib to TACE. At best, the concomitant use of sorafenib might modestly delay, but not prevent, tumor progression after TACE. Larger, well-designed phase III trials with a survival endpoint are needed before it can be concluded that there is any benefit for the addition of sorafenib to TACE.
Sorafenib is a multikinase inhibitor acting on the vascular endothelial growth factor receptor, among others. Findings from the SHARP trial, which showed that sorafenib significantly prolonged survival compared with supportive care alone in patients with advanced HCC, established sorafenib monotherapy as a new reference standard for systemic treatment of advanced HCC and formed the basis for approval of sorafenib for unresectable HCC in the United States. (See "Systemic treatment for advanced hepatocellular carcinoma", section on 'Sorafenib'.)
However, the benefit of combined therapy over TACE alone has been directly addressed in three randomized phase II trials and two phase III trials, only one of which, presented in abstract form only, suggests benefit:
●A lack of benefit for post-TACE sorafenib was shown in an initial phase III trial in which 458 patients with unresectable HCC, Child-Turcotte-Pugh class A cirrhosis, and ≥25 percent tumor necrosis/shrinkage after one or two TACE sessions were randomly assigned to sorafenib (400 mg twice daily) or placebo until progression/recurrence or unacceptable toxicity . The difference in median time to progression in the sorafenib group was not statistically significant (5.4 versus 3.7 months, HR for progression 0.87, 95% CI 0.7-1.09), and there was no difference in overall survival. The median daily dose of sorafenib was only 386 mg, and the median time of administration was only 17 weeks. These factors may have compromised the ability to detect benefit.
●Similarly, there was no benefit for sorafenib in a European phase III trial in which 313 patients with unresectable liver-confined HCC were randomly assigned to sorafenib (400 mg twice daily) or placebo; all patients underwent a single DEB-TACE treatment via the hepatic artery two to five weeks following randomization . The trial was stopped prematurely for futility. There was no difference in median progression-free survival (PFS) in the sorafenib and placebo groups (median 238 versus 235 days, respectively, HR 0.99). The median daily sorafenib dose was 660 mg, and the median duration was 120 days.
●Two other randomized phase II trials also failed to show clear benefit from the addition of sorafenib, although as with the two phase III trials described above, the duration of sorafenib treatment was short (17 and 21 weeks, respectively) [168,169].
●The only positive randomized phase II trial, the TACTICS trial, randomly assigned 156 Japanese patients with unresectable HCC to TACE alone or to sorafenib 400 mg once daily for two to three weeks prior to TACE and then 800 mg twice daily after TACE was started . In the latest analysis, median PFS, the primary endpoint, was significantly higher with combined therapy (22.8 versus 13.5 months, HR for progression 0.66, 95% CI 0.47-0.94), but the difference in median overall survival was not statistically significant (36.2 versus 30.8 months, HR 0.86, 95% CI 0.61-1.22) . Notably, patients were treated with sorafenib until progression to the point where further TACE treatments were not feasible, and the median duration of sorafenib therapy was 38.7 weeks. The authors postulated that the longer duration of therapy may have contributed to the positive result of this study compared with the other four.
Taken together, these results (and the conclusions of a meta-analysis of three of the published trials [166,168,169], and several other nonrandomized prospective and retrospective reports ) suggest that at best, the concomitant use of sorafenib might modestly delay, but not prevent, tumor progression after TACE. Larger, well-designed phase III trials with a survival endpoint are needed before it can be concluded that there is any benefit for the addition of sorafenib to TACE.
Embolization versus systemic therapy — Given the more favorable side effect profile, locoregional forms of therapy, such as TACE and radioembolization, may be preferable to systemic therapy alone for initial treatment of patients with locally advanced unresectable HCC without extrahepatic metastases, as long as they are suitable candidates. However, the only available data compare embolization with sorafenib monotherapy; there are no data comparing embolization with bevacizumab plus atezolizumab, which is currently preferred over sorafenib alone because of greater survival and higher response rates. Such trials are ongoing. (See "Systemic treatment for advanced hepatocellular carcinoma".)
Radioembolization versus sorafenib — Systemic therapy with the multitargeted agent sorafenib is one first-line treatment option for patients with advanced unresectable HCC; it offers the potential for prolonged survival over supportive care alone. (See "Systemic treatment for advanced hepatocellular carcinoma", section on 'Sorafenib'.)
Whether results with radioembolization are better than those that can be achieved with sorafenib was directly addressed in two different multicenter trials, both of which concluded that outcomes were similar:
●In the multicenter phase III randomized SIRveNIB trial, 360 Asia-Pacific patients with newly diagnosed unresectable HCC without portal vein thrombus or extrahepatic metastases (89 percent Child-Turcotte-Pugh class A cirrhosis (table 1)) were randomly assigned to sorafenib (400 mg twice daily) or a single injection of 90Y microspheres . 90Y was associated with a significantly higher tumor response rate (16.5 versus 1.7 percent), fewer adverse events (constipation, diarrhea, fatigue, rash, hand-foot skin reaction, hypertension), and fewer serious adverse events (at least one treatment-related serious adverse event in 4.6 versus 9.3 percent) compared with sorafenib. The overall disease control rate (objective response plus stable disease; 41.8 versus 42.7 percent), time to tumor progression (5.88 versus 5.36 months), and median overall survival (8.8 versus 10 months) were not significantly different.
●Similar results were noted in a second trial conducted in France in which 467 patients with no extrahepatic metastases (83 percent Child-Turcotte-Pugh class A cirrhosis) and new HCC not eligible for resection, transplantation, or thermal ablation; previously cured HCC (after resection or thermal ablation); or HCC with two unsuccessful rounds of TACE were randomly assigned to radioembolization or sorafenib . 90Y was associated with a modestly higher objective tumor response rate (19 versus 12 percent), a modestly lower overall disease control rate (68 versus 78 percent), fewer adverse events (fatigue, diarrhea, hand-foot skin reaction), and similar median survival (8 versus 9.9 months, HR 1.15, 95% CI 0.94-1.41).
It should be noted that neither of these trials were a noninferiority trial, and both failed to meet their primary outcome endpoint (overall survival difference). Although these two trials were "negative" in that regard, given the more favorable adverse event profile, locoregional forms of therapy, such as radioembolization, may be preferable to systemic therapy alone for initial treatment of patients with locally advanced unresectable HCC without extrahepatic metastases, as long as they are suitable candidates.
Benefit could also not be shown for the addition of radioembolization to sorafenib in the phase III Sorafenib and Micro-therapy Guided by Primovist Enhanced MRI in Patients with Inoperable Liver Cancer (SORAMIC) trial, conducted in 424 patients who were not eligible for TACE (90 percent had Child-Turcotte-Pugh class A cirrhosis) . The addition of radioembolization to sorafenib did not result in a significant improvement in overall survival compared with sorafenib alone (median 12.1 versus 11.4 months), and combined therapy was associated with higher rates of grade 3 or 4 adverse events.
A year 2020 Cochrane analysis concluded that the evidence on the effects of radioembolization with or without sorafenib compared with sorafenib alone was highly insufficient . Radioembolization versus sorafenib seemed to achieve similar survival and to cause fewer adverse effects, but the quality of the evidence was low. It could not be determined if the addition of radioembolization to sorafenib affected all-cause mortality or the occurrence of adverse effects.
TACE versus sorafenib — In addition, a third trial that tested whether the combination of TACE plus RT was better than initial sorafenib in patients with liver-isolated HCC with macrovascular invasion also concluded that first-line treatment with TACE plus RT was better tolerated and provided improved PFS, objective response rates, time to progression, and overall survival compared with sorafenib treatment. This trial is discussed in detail below. (See 'TACE plus radiation therapy' below.)
Embolization versus bevacizumab plus atezolizumab — There are no trials comparing any form of embolization with bevacizumab plus atezolizumab, which, in the IM Brave 150 trial, provided a significantly better survival and response rate compared with sorafenib monotherapy. (See "Systemic treatment for advanced hepatocellular carcinoma".)
An important point that may influence the choice of initial local therapy (eg, embolization) versus bevacizumab plus atezolizumab is whether initial local therapy alters the subsequent response to bevacizumab plus atezolizumab. The only available data come from a unplanned subgroups analysis of the IM Brave 150 trial, in which approximately one-half of the patients had received initial local therapy (not specified). In unplanned subgroup analysis, the HR for survival for combined therapy over sorafenib monotherapy was 0.57 (95% CI 0.38-0.87) in those without prior local treatment, and 0.68 (95% CI 0.39-1.01) in those with prior local treatment. These results should be viewed as hypothesis-generating only.
Hepatic intraarterial chemotherapy without embolization — Where the technical expertise is available, hepatic arterial infusion chemotherapy (HAIC) is an alternative to TACE for patients with large unresectable HCCs. However, until further data are available, we do not consider HAIC to be a preferred approach.
TACE has been a standard approach for treatment of large unresectable HCCs. However, some have shown that the use of multiagent chemotherapy plays an important role in improving survival in advanced unresectable HCC while the addition of particle embolization increases the frequency of adverse effects but does not improve survival . Furthermore, it is often difficult to perform a complete embolization of large HCCs because of plentiful extrahepatic collateral arteries. This has led to interest in HAIC using multiagent chemotherapy, which provides sustained local high concentrations of chemotherapy agents in tumors, without the need for embolization.
The benefit of this strategy over repeated courses of TACE alone was shown in a phase III trial in which 315 patients with unresectable large (≥7 cm) HCC without macrovascular invasion or extrahepatic spread were randomly assigned to HAIC via the hepatic artery using FOLFOX (oxaliplatin 130 mg/m2 plus fluorouracil [FU] bolus 400 mg/m2 on day 1, and followed by FU infusion 2400 mg/m2 for 24 hours, once every three weeks for up to six courses) versus superselective TACE (using epirubicin, lobaplatin, and lipiodol and polyvinyl alcohol particles) repeated every six weeks .The entire patient population was Asian, and had predominantly Child-Turcotte-Pugh class A cirrhosis (81 percent) and HBV as the underlying HCC risk factor (89 percent). The benefits of HAIC over TACE included significantly better median overall survival (23.1 versus 16.1 months; HR 0.58, 95% CI 0.45-0.75), a higher objective response rate (46 versus 18 percent), a longer median PFS (9.6 versus 5.4 months), and better tolerability (incidence of serious adverse effects during treatment 19 versus 30 percent). However, abdominal pain was observed in 37 of the 157 patients receiving HAIC when oxaliplatin was injected, and it resolved when the injection was stopped. Furthermore, 12 patients had thrombosis or dislocation of the catheter tip in this group, and required catheter revision, and four developed gastric ulcers during therapy, one of which had upper GI bleeding that required endoscopic hemostasis.
There are several concerns about this study which renders interpretation of the results difficult:
●The study was open label, and there were significant differences in poststudy treatment that may have been influenced by investigator decisions, and affected the overall survival outcomes. As an example, more patients in the HAIC group underwent subsequent hepatic resection (38 versus 18 in the TACE group). Crossover to the alternative treatment was much more frequent in the TACE group (20 versus 8 patients).
●One of the agents used in the TACE procedure, lobaplatin, is only available in China.
●The median number of treated cycles for HAIC was 4 (range 2 to 5), and only 28 of the 157 patients in the HAIC group received the planned six cycles of treatment. The median number of TACE procedures was only 2 (range 1 to 3).
●Eligibility was limited to patients with either no cirrhosis or cirrhosis of no worse than Child-Turcotte-Pugh class A. However, the number of patients without cirrhosis was not reported. Whether the results are generalizable to a broader population, including those with cirrhosis or HCC etiologies other than HBV, is uncertain.
Furthermore, HAIC is expensive, requires specific technical expertise and an invasive procedure (percutaneous access of the hepatic artery) every three weeks, and inpatient management of the chemotherapy infusion. For all of these reasons, we consider that these data are not practice changing, and we await additional studies before concluding that HAIC is a preferred approach over TACE for treatment of large unresectable HCCs.
Stereotactic body radiation therapy — RT techniques such as stereotactic body radiotherapy (SBRT) are reasonable options for nonsurgically managed patients being considered for other liver-directed therapies who have no extrahepatic disease, have limited tumor burden, and have relatively preserved liver function. The choice of SBRT over other liver directed therapies generally depends on institutional expertise and patient preference (SBRT generally requires several visits to plan and execute).
SBRT is a technique in which a single (sometimes called stereotactic radiosurgery) or a limited number of high-dose RT fractions (typically three to six) are delivered to a small, precisely defined target through the use of multiple nonparallel radiation beams. The beams converge on the target lesion, minimizing radiation exposure to the adjacent normal tissue. This targeting allows treatment in either a single or a limited number of dose fractions.
Experience with SBRT for primary liver tumors is increasing. A wide variety of dosing and fractionation schedules have been used, with total doses ranging from 24 to 60 Gy over three to six fractions. Overall, local tumor control rates two to three years following SBRT for HCC range from 68 to 95 percent [177-188]. However, in contrast to the encouraging two- and three-year local control rates, two- to three-year PFS rates (which range from 21 to 48 percent) and overall survival rates (which range from 21 to 69 percent) are lower due to out-of-field progression in the untreated parts of the liver [177,180-184,186-192].
●In one of the largest retrospective series, 93 patients (69 Child-Turcotte-Pugh class A cirrhosis, the rest Child-Turcotte-Pugh class B cirrhosis) with a small HCC 1 to 6 cm in diameter (median 2 cm) who were not suitable for surgical resection were treated with SBRT . The in-field complete response rate was only 16 percent, but the in-field PFS rate at three years was 92 percent. Most local failures developed in patients with tumor size >3 cm, in whom the three-year local control rate was 76 percent; the corresponding rates for patients with tumor size 2.1 to 3 cm and ≤2 cm were 93 and 100 percent, respectively. Out-of-field intrahepatic recurrence-free survival rates at one and three years were 52 and 32 percent, respectively. Hepatotoxicity grade ≥3 occurred in only six patients (6.5 percent).
●A multicenter phase II trial of 74 patients with unresectable HCC (median 2.4 cm, range, 1.1 to 9.9) treated with SBRT (45 to 60 Gy in three fractions) reported two- and three-year local control rates of 97 and 95 percent, respectively, and two- and three-year PFS rates of 48 and 36 percent, respectively. There were only three grade 3 or worse acute liver toxicities within three months of treatment completion.
The place of SBRT in the therapeutic armamentarium for unresectable HCC not eligible for transplant or local thermal ablation is uncertain. There are no prospective data comparing SBRT with other locoregional therapies. Single-institution data from the University of Michigan show comparable efficacy and safety compared with RFA in these patients; for tumors larger than 2 cm, SBRT may provide more effective local control with a lower complication rate . One analysis from the National Cancer Database suggests that treatment with RFA has a higher survival rate compared with SBRT; however, there is significant confounding selection bias in the data used for this study as many important factors that affect treatment outcome were not accounted for in this database (RFA tended to be used for smaller, well-encapsulated, and more easily delineated lesions away from other organs and blood vessels and in healthier patients) .
SBRT has been generally well tolerated; reported severe toxicity rates are generally less than 10 percent in patients with Child-Turcotte-Pugh class A disease [112,177,180-184,195]. Over the past two decades, a greater understanding of the dose-volume effects of partial-liver RT has enabled radiation therapists to predict the risk of classic RILD associated with a given treatment plan using normal tissue complication probability models [196,197].
As with other nonsurgical ablative therapies, response assessment following SBRT for HCC is difficult. (See 'Assessing response to locoregional therapies' above.)
Although 37 to 85 percent of lesions demonstrate a partial or complete response using Response Evaluation Criteria in Solid Tumors (table 2) and other conventional response assessment algorithms [177,178,180-182,189,191,195], few studies have examined the relationship between SBRT and other radiographic response criteria based upon visualization of tumor necrosis and/or intratumoral arterial enhancement, as are used to predict outcomes following other treatments for HCC, such as TACE or RFA (ie, LI-RADS] treatment response algorithm (algorithm 2)). (See 'Assessing response to locoregional therapies' above.)
Complicating matters further, patients treated with SBRT may have a delayed response to therapy on dynamic cross-sectional imaging. Issues related to response assessment in HCC are discussed in detail elsewhere. (See "Assessment of tumor response in patients receiving systemic and nonsurgical locoregional treatment of hepatocellular cancer", section on 'Classification systems that incorporate tumor viability'.)
Indications — Consistent with the dearth of comparative trials, there is a general lack of consensus as to the appropriate indications for any form of RT in patients with HCC:
●2018 clinical practice guidelines from the AASLD do not address the utility of any form of RT alone for treatment of HCC .
●Consensus-based guidelines from the NCCN list RT (conformal or stereotactic) as an alternative option to ablation or arterially directed therapies for patients with unresectable HCC who are not transplantation candidates .
●An expert consensus group of the AHPBA concluded that RT can provide local control for some unresectable HCC lesions, that better RT planning and delivery (eg, hypofractionation, stereotactic treatment, and proton beam therapy) has advanced the ability to escalate the radiation dose to unresectable HCCs without causing undue toxicity, and that strategies combining RT with other therapies merit continued evaluation . However, they did not make specific recommendations as to which patients should be considered for any RT strategy.
In our view, SBRT is a reasonable option for selected patients who are being considered for other local treatment modalities and have no extrahepatic disease, have limited tumor burden (ie, the radiation treatment portal will be small), and have relatively preserved liver function. In fact, some centers utilize SBRT for potentially transplantable tumors as "bridging" therapy. There is currently an ongoing international clinical trial comparing TACE-DEB with SBRT prior to transplantation, evaluating differences in toxicity, outcome, and histologic changes in the explant specimens. (See "Liver transplantation for hepatocellular carcinoma", section on 'Stereotactic radiotherapy'.)
Contraindications — Use of RT should be limited to patients with well-compensated liver function (Child-Turcotte-Pugh class A, or early Child-Turcotte-Pugh class B with a score of 7 (table 1)) who have adequate liver volume outside of the radiation field. Child-Turcotte-Pugh class B patients with a score of ≥8 or Child-Turcotte-Pugh class C patients ought to be treated in a clinical trial or in the bridge-to-transplant setting given the elevated risk of radiation-induced hepatic toxicity in these patients .
For patients with prior hepatic RT, it is recommended that these patients be evaluated in a tertiary care center with expertise in hepatic RT when considering reirradiation [199,200].
Complications — Minimizing radiation-induced complications depends on careful patient selection and RT treatment planning. Common acute effects include transient fatigue and nausea. Possible long-term effects of radiation to the liver include worsening hepatic function, with associated ascites, hepatomegaly, thrombocytopenia, and elevated liver function tests [201-205]. Rarely, cases of radiation-induced biliary stenosis or death from radiation-induced liver failure have been reported [206,207]. For lesions near the capsule of the liver, there can be transient right upper quadrant pain associated with stretching of the capsule from swelling after treatment. Additionally in situation where SBRT is delivered to the periphery of the liver, care must be taken to minimize radiation to adjacent organs to prevent toxicity (such as radiation pneumonitis, rib fracture, chest-wall pain, bowel or stomach ulceration, or cardiac effects).
PATIENTS WITH PORTAL VEIN INVASION OR THROMBUS — For patients with invasion or tumor thrombus involving the portal vein, options include radioembolization, stereotactic body radiation therapy (SBRT), transarterial chemoembolization (TACE) plus radiation therapy (RT), proton beam irradiation (where available), or initial systemic therapy. Surgery may be an option if the patient is otherwise resectable, especially if the portal vein thrombus is limited to a first-order branch . (See "Surgical management of potentially resectable hepatocellular carcinoma".)
For patients who are deemed not to be surgical candidates, based on a single trial that demonstrated that TACE plus RT provides outcomes that are better than those that can be achieved with sorafenib alone , TACE plus RT is a reasonable option for patients with tumor invasion of a first or second branch of the portal vein as long as unilateral portal blood flow is preserved, especially if they have hepatitis B virus (HBV) related HCC. (See 'TACE plus radiation therapy' below.)
For patients with tumor thrombus involving the main portal vein or a major branch, with either no portal vein flow or reversal of portal vein flow, TACE should be avoided. (See 'Patient selection and contraindications' above.)
Other options, such as SBRT, radioembolization, sorafenib plus hepatic arterial infusion of chemotherapy (HAIC; where available), proton beam irradiation (where available), or systemic therapy with atezolizumab plus bevacizumab are more appropriate in this setting. (See "Systemic treatment for advanced hepatocellular carcinoma".)
Portal vein thrombus (especially involvement of the main portal vein) is a poor prognostic factor in patients with unresectable HCC. In general, survival is short, and patients who are initially treated with systemic therapy (sorafenib) are prone to rapid disease progression and early death [9,210-213].
Results may be better with radioembolization, proton beam irradiation, SBRT, or, in the absence of tumor thrombus involving the main portal vein or a major branch, with either no portal vein flow or reversal of portal vein flow, TACE plus radiotherapy.
Radioembolization — While main portal vein thrombus has historically been considered a contraindication to radioembolization, collective experience in over 200 patients with main portal vein thrombus suggests the safety and efficacy of this approach [135,136,214,215].
There is theoretically less arterial ischemia induced by radioembolization because of the particle size (32 versus 70 to 300 microns with beads), suggesting that it should be safer in the setting of portal vein thrombus. Compared with TACE, rates of severe adverse effects with radioembolization appear low . However, although radioembolization appears to be safe in these patients, median survival is short in many series, particularly in those with Child-Turcotte-Pugh class B cirrhosis. (See 'Radioembolization' above.)
Proton beam irradiation — Charged particle irradiation (ie, proton beam and carbon ion RT) has been used in a limited number of patients, especially in Japan [217,218]. Where available, proton beam irradiation is a reasonable approach for patients with a large HCC and associated portal vein thrombus, although SBRT may also be helpful in this situation. There is a growing body of evidence, primarily from Japan, supporting the use of proton beam irradiation, particularly for patients with large tumors or portal vein thrombus [217-226] (see "Radiation therapy techniques in cancer treatment", section on 'Particle therapy'). While proton therapy may be helpful for some patients, it is generally not available for most patients; there are only 27 proton facilities currently operational in the United States.
●In one study, 162 patients with 192 HCCs were treated with proton beam irradiation . Most tumors had a diameter of 3 to 5 cm. The majority of the patients had received some form of previous nonsurgical therapy. The five-year local control rate was 87 percent. Overall survival at five years was 24 percent, with most patients dying from metachronous intrahepatic lesions. Acute and late gastrointestinal toxicities were uncommon.
●The utility of proton beam therapy in the setting of portal vein thrombus was shown in a report of 12 patients who had HCC and tumor thrombus in the main trunk or major branches of the portal vein . The objective response rate was 100 percent, a far higher value than that reported in previous studies of patients with portal vein thrombus, and five recanalized their portal vein.
●In another report of 44 patients with HCC (15 with tumor vascular thrombus) treated with hypofractionated proton beam irradiation (15 fractions of proton therapy to a maximum total dose of 67.5 Gy equivalent), the two-year local control rate was 95 percent, although the majority progressed in other sites (two-year progression-free survival 40 percent) .
Stereotactic body radiation therapy — SBRT may be complementary to other existing ablative options in patients with tumors that are otherwise not amenable to other local treatment modalities, such as those with tumor vascular thrombus. This was shown in a prospective evaluation of 102 patients with HCC who were not considered candidates for surgical resection, TACE, or radiofrequency ablation and were enrolled in sequential phase I and II trials of SBRT [104,178]. The median prescription dose was 36 Gy in six fractions. These were generally patients with advanced disease; 51 percent had a Cancer of the Liver Italian Program score of 2 to 4 (table 3), 61 percent had multiple lesions, the median diameter of the largest lesion was 7.2 cm, and 55 percent had tumor vascular thrombus. Despite these adverse features, the one-year local control rate was 87 percent, and the median time to local progression had not been reached at a median follow-up of 31.4 months. Although responses were observed in patients with tumor vascular thrombus, the fraction (response rate) was not reported. The one-year survival rate was 55 percent.
Whether results with SBRT are better than those that can be achieved with radioembolization in this setting is unclear; there are no randomized trials. A meta-analysis of 37 studies of radioembolization, SBRT, or three-dimensional conformal radiation therapy (3D-CRT; all observational or single-arm prospective studies) concluded that while overall response and local control rates were different (and for the most part, favored SBRT over radioembolization), the response rate in the portal vein thrombus was similar (39 versus 35 percent), as was overall survival at both one and two years .
TACE plus radiation therapy — Another approach for patients with HCC and portal vein thrombus not involving the main portal vein or associated with reduced portal blood flow is combined therapy with TACE plus RT [106,228-231]. In a registry database series of 412 patients with unresectable HCC complicated by thrombus in the portal vein (49 percent bilateral or main portal vein) who received focal 3D-CRT combined with TACE before or after RT (lipiodol, cisplatin gelatin sponge), 40 percent had an objective response in the portal vein tumor thrombus, and 43 percent were still alive at one year; however, grade 3 or 4 hepatotoxicity was seen in 10 percent of patients during or within three months of completion of RT .
Results with combined therapy may be even better using newer planning methods for RT (eg, 3D-CRT or hypofractionated SBRT), with which objective response rates with RT alone directed to the portal vein thrombus range from 39 to 62 percent [105,107,108,110-112,228,232,233].
There are no trials directly comparing TACE plus RT with other forms of locoregional therapy. One trial of TACE plus RT delivered using 3D-CRT versus sorafenib for patients with macroscopic vascular invasion compared TACE plus RT (within three weeks of the first TACE, 45 Gy using 3D-CRT planning with a fraction size of 2.5 to 3 Gy) with initial sorafenib (400 mg twice daily) in 90 treatment-naive patients with liver-confined HCC invading the first or second branch of the portal vein with preserved unilateral portal blood flow (ie, no main portal vein thrombus) . All patients had portal vein invasion and Child-Turcotte-Pugh class A liver function. The TACE procedure was repeated every six weeks for the first six months and every six to eight weeks thereafter. The group receiving TACE plus RT had a significantly higher progression-free survival rate at 12 weeks (the main endpoint, 87 versus 34 percent), a higher radiographic response rate at 24 weeks (33 versus 2 percent), significantly longer median time to progression (31 versus 12 weeks), and significantly greater median overall survival (55 versus 43 weeks). Potentially curative resection could be performed on five patients between 27 and 40 weeks after beginning TACE and RT; four remained alive at last follow-up with an overall survival time of 119 to 149 weeks. Notably, this study was conducted in an Eastern, HBV-predominant population; the results require validation in Western populations.
Sorafenib plus hepatic arterial infusion of chemotherapy — Another approach, where available, is sorafenib plus HAIC. Cisplatin-based HAIC is widely used in Japan as an alternative to sorafenib for patients with HCC and portal vein invasion . Although this approach is associated with higher response rates than those that are achieved using systemic chemotherapy alone or sorafenib monotherapy, questions have persisted about whether there is a survival benefit [235-238].
This was addressed in a randomized trial directly comparing sorafenib alone (400 mg twice daily) versus sorafenib plus oxaliplatin-based HAIC (oxaliplatin 85 mg/m2, leucovorin 400 mg/m2, and fluorouracil bolus 400 mg/m2, all on day 1, followed by fluorouracil infusion 2400 mg/m2 for 46 hours, every three weeks, given by repeated catheterization rather than a fixed intra-arterial catheter) in 247 Chinese patients with HCC (81 percent HBV related) and portal vein invasion . The median age was 49, and approximately 37 percent had tumor invasion of the main portal vein. Median overall survival, the primary endpoint, was significantly higher for combined therapy (13.37 versus 7.13 months, hazard ratio [HR] 0.35, 95% CI 0.26-0.48). Combined therapy was associated with a higher rate of grade 3 or 4 neutropenia, thrombocytopenia, and vomiting. Whether these results can be extrapolated beyond China (where patients with HCC tend to be young with predominant HBV infection, as was seen in this trial) is unclear. Furthermore, whether these results are better than those that can be achieved using sorafenib plus systemic chemotherapy with an oxaliplatin-containing regimen or atezolizumab plus bevacizumab is also unclear . Randomized trials are needed. (See "Systemic treatment for advanced hepatocellular carcinoma", section on 'Sorafenib plus chemotherapy'.)
SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The COVID-19 pandemic has increased the complexity of cancer care. Important issues in areas where viral transmission rates are persistently high include balancing the risk from delaying diagnostic evaluation and cancer treatment versus harm from COVID-19, minimizing the number of clinic and hospital visits to reduce exposure whenever possible, mitigating the negative impacts of social distancing on delivery of care, and appropriately and fairly allocating limited health care resources. Specific guidance for decision-making for treatment of HCC during the COVID-19 pandemic is available from the International Liver Cancer Association (ILCA). General recommendations for cancer care during active phases of the COVID-19 pandemic are discussed separately. (See "COVID-19: Considerations in patients with cancer".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Hepatocellular carcinoma".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Liver cancer (The Basics)")
SUMMARY AND RECOMMENDATIONS
•Hepatocellular carcinoma (HCC) is an aggressive tumor that frequently occurs in the setting of cirrhosis. Treatment options are divided into potentially curative surgical therapies (ie, resection, liver transplantation) and nonsurgical therapies which can be liver-directed, or systemic. (See 'Introduction' above.)
Disease extent and hepatic reserve generally dictate the therapeutic approach. An algorithmic approach to treatment is useful to conceptualize the various treatment options that are available, but it may not be applicable in all settings (algorithm 1). Alternative algorithms are available such as the one that is used by the Barcelona Clinic Liver Cancer group; their most recent updated algorithm is available online. (See 'Treatment algorithms and general approach to the patient with localized disease' above.)
In general, for suitable candidates, locoregional liver-directed therapies are generally preferred over systemic therapy for initial treatment of unresectable HCC without extrahepatic metastases, given the more favorable side effect profile. Embolization and HAIC are generally reserved for those who are not eligible for local thermal ablation or focal stereotactic body radiotherapy (SBRT). (See 'Embolization versus systemic therapy' above and "Overview of treatment approaches for hepatocellular carcinoma", section on 'Unresectable disease'.)
•All patients undergoing therapy for HCC should have proper monitoring and treatment of their underlying liver disease. (See 'Importance of multidisciplinary care' above.)
•Accurate response assessment for many nonsurgical locoregional therapies requires evaluation of posttreatment residual tumor enhancement, as assessed by the Liver Reporting and Data System (LI-RADS) treatment response algorithm (algorithm 2). In some cases, serial assay of tumor markers such as alpha fetoprotein may also be used. (See 'Assessing response to locoregional therapies' above.)
●Patients without portal vein thrombus
•Hepatic arterial embolization is an appropriate option for patients with a large unresectable or multifocal HCC who have no worse than Child-Turcotte-Pugh class A or B cirrhosis (table 1) and no extrahepatic tumor spread, vascular invasion, or tumor thrombus involving the main portal vein or one of its lobar branches. It is also appropriate for "bridging" therapy in a patient awaiting liver transplantation and for selected patients prior to resection of a large HCC, particularly involving the right lobe, in conjunction with portal vein embolization.(See 'Embolization' above.)
•For most patients in whom hepatic arterial embolization is indicated, we suggest transarterial chemoembolization (TACE) or transarterial radioembolization (TARE) rather than bland embolization (Grade 2C). If TACE is chosen, we suggest TACE with drug-eluting beads (DEB-TACE) over conventional TACE given the more favorable side effect profile (Grade 2C). However, we restrict TACE to patients with a good performance status (table 4), and no vascular invasion, main or lobar branch portal vein thrombus, extrahepatic disease spread, encephalopathy, or biliary obstruction. (See 'Is there an optimal embolization approach?' above and 'Drug-eluting beads' above.)
For patients with intermediate-sized HCC (ie, 3 to 5 cm), we suggest TACE plus radiofrequency ablation or microwave ablation rather than TACE alone (Grade 2C). (See 'TACE plus local treatments' above.)
•For patients undergoing DEB-TACE, we generally repeat the procedure for each lesion, one month apart. Otherwise, if other techniques are used (eg, conventional TACE), we do not repeat the TACE procedure unless there is clear evidence of tumor progression or residual viable tumor. The decision to retreat with TACE for clearcut disease progression must be individualized and based upon an assessment of tumor response and the hepatic reserve (See 'Retreatment' above.)
•Although there is little consensus on this issue, we generally choose TARE over TACE for patients with unresectable HCC who are otherwise adequate candidates for local embolization therapy but who have macrovascular invasion, including portal vein thrombus. Where available, and for appropriately selected patients, superselective radioembolization (segmental radioembolization, also called radiation segmentectomy) may provide high rates of local control with less radiation-induced liver disease. (See 'Radioembolization' above.)
•Where the technical expertise is available, hepatic intra-arterial chemotherapy (HAIC) is an alternative to TACE for patients with large unresectable HCCs. However, HAIC is expensive, requires specific technical expertise, and inpatient management every three weeks. We await additional studies before concluding that HAIC is a preferred approach over TACE for treatment of large unresectable HCC. (See 'Hepatic intraarterial chemotherapy without embolization' above.)
•RT techniques such as SBRT are reasonable options for nonsurgically managed patients being considered for other liver-directed therapies who have no extrahepatic disease, have limited tumor burden, and have relatively preserved liver function. The choice of SBRT over other liver directed therapies generally depends on institutional expertise and patient preference. (See 'Stereotactic body radiation therapy' above.)
●Patients with portal vein thrombus
•For patients with tumor thrombus involving the portal vein, options include TARE, SBRT, TACE plus RT, proton beam irradiation, or initial systemic therapy. For patients with tumor invasion of a first or second branch of the portal vein who have preserved unilateral portal blood flow we suggest TACE plus RT rather than systemic therapy (Grade 2B). (See 'Patients with portal vein invasion or thrombus' above.)
•For patients with tumor thrombus involving the main portal vein or a major branch, with either no portal vein flow or reversal of portal vein flow, TACE is contraindicated, and other options, including atezolizumab plus bevacizumab, are more appropriate.