Embolotherapy for Hepatic Oncology: Current Perspectives and Future Directions

Introduction

Primary and secondary liver cancers represent a significant cause of cancer-related death worldwide. Primary liver cancer, including hepatocellular carcinoma and cholangiocarcinoma, is the sixth most common cancer worldwide and the fourth most common cause of cancer death¹. Secondary, or metastatic, liver cancer is more common than primary liver cancer, with colorectal liver metastases being the most common². Other common metastases are from gastro-entero-pancreatic neuroendocrine tumors (GEP-NETs), breast, lung, and melanoma.

While surgical intervention is often considered “curative,” only a minority of patients with primary and secondary liver tumors are surgical candidates^3–6. Thus, trans-arterial embolotherapeutic techniques represent an important therapeutic option in hepatic oncology. Approximately 75% of the blood supply to normal liver parenchyma is supplied by the portal vein, and the remaining 25% by the hepatic artery. Liver tumors, on the other hand, are almost exclusively supplied by the hepatic artery. This pathophysiologic phenomenon is the underlying principle exploited by transarterial hepatic embolotherapy^7,8. Embolotherapy is a transarterial catheter-based locoregional therapy whereby any of a variety of therapeutic embolic agents are delivered via the hepatic artery, interrupting the tumoral arterial blood supply while delivering the therapeutic agent preferentially to the areas of tumor involvement. The tumor becomes hypoxic, ischemic, and subsequently undergoes coagulative necrosis. Embolotherapy plays an important role in the management of both primary and secondary liver tumors, ranging from curative as a standalone therapy, an adjunct to additional therapies, or as a salvage/palliative option. The objectives of this review article are to explain the concepts behind hepatic embolotherapy for primary and secondary liver cancer, and to summarize the most current available literature reporting the indications, safety, and efficacy of embolotherapy in hepatic oncology.

Technical Overview and Patient Selection

Technical Overview

Hepatic embolotherapy consists of three primary modalities: transarterial embolization (TAE), transarterial chemoembolization (TACE), and transarterial radioembolization (TARE) or selective internal radiotherapy (SIRT). The technical aspects of each modality are detailed in Table 1. These transarterial therapies are performed after accessing either common femoral artery or the left radial artery. After placement of a vascular sheath in the site of entry, hepatic angiography is then performed, often with the assistance of intra-procedural cone-beam computed tomography (CT) and post-processing software to delineate tumor-feeding vessels of interest. Once the artery supplying tumor is catheterized, known as selective or super-selective catheterization depending on the location within the hepatic arterial system, embolotherapeutic agents are delivered through the catheter, interrupting the tumoral blood supply to the desired angiographic endpoint. The ideal catheter position depends in large part by the transarterial therapeutic modality, but is generally recommended to be as selective as possible without compromising access to tumor supplying arteries, while minimizing exposure to normal hepatic parenchyma.

Table 1.

Overview of Hepatic Embolotherapeutic Techniques

Embolotherapy technique	Mechanism	Advantages	Disadvantages
Transarterial embolisation (TAE) or ‘bland embolization’	Permanent occlusion of the tumor vascular bed - require small particles (50 – 120 μn), such as polyvinyl alcohol (PVA) particles or microspheres, made of trisacryl gelatin, are used 9–10.	Relatively inexpensive	Marked post-embolization syndrome, potentially necessitating hospital admission
		Readily available “off the shelf”
		No chemotherapy induced hepatocellular toxicity
			Small particles may shunt to the lungs, causing pulmonary infarction
Conventional transarterial chemoembolization or (cTACE)	Two-step procedure: Embolic material loaded with chemotherapy drug delivered in lipiodol emulsion. Typical chemotherapy agents used are doxorubicin, cisplatin, mitomycin C, irinotecan or gemcitabine, alone or in combination 11-13. Bland embohzation with temporary agent e.g. Gelfoam®14	Deliver higher doses of cytotoxic chemotherapy and potentiate therapeutic action, while decreasing the systemic effects of the chemotherapy agent15,16	Post-embolization syndrome
		Highest level of evidence in treatment of HCC and incorporated in BCLC* staging system
			Chemotherapy drug still reaches systemic circulation
			No standardized protocol
Drug-eluting bead transarterial chemoembolization (DEB-TACE)	Specialized microspheres are loaded with antitumor drugs, most commonly doxorubicin or irinotecan, which are then delivered transarterially to the tumor.	Drug delivery more predictable and sustained, with higher intra-tumoral concentrations and lower peak systemic doses of chemotherapy compared to cTACE 17,18.	Post-embolization syndrome
			More expensive than cTACE
		DEB irinotecan TACE more effective in treating colorectal metastases than cTACE19	Potential biliary necrosis/risk of biloma formation in treating neuroendocrine tumors 20
Selective internal radiation therapy (SIRT)	Type of intra-arterial brachytherapy: Microspheres containing the radioactive element yttrium-90 (Y-90) cause DNA-induced damage and radiation necrosis. Two FDA approved devices: Glass microspheres (Therasphere): FDA approved as a palliative option for advanced HCC Resin microspheres: FDA approved for progressive colorectal liver metastases (CLM)	Deliver high doses of radiation to tumor and minimize dose to radiation sensitive liver	Most expensive of all techniques
			Requires institutional development of service, with coordination between multiple departments and nuclear regulatory commission (NRC) Radiation adverse events to patient; such as radiation-induced liver disease (REILD), biliary toxicity, cirrhosis, and small particles may shunt to the lungs, causing pulmonary fibrosis
		Role in all stages of HCC; curative, downstage to transplant eligibility, local control and possibly safer in advanced disease (portal vein invasion)
	Two types of specialized dosimetric techniques over standard delivery:
	Radiation segmentectomy – ‘curative’ therapy by treating ≤ two hepatic segments using a lobar dose Radiation lobectomy – local tumor control and contralateral lobar hypertrophy

Currently, the optimal endpoint of TACE is not formally established by current evidence, and embolization is continued until achievement of the chosen fluoroscopic endpoint: either stasis or sub-stasis (sub-stasis is defined as contrast clearance within 2 – 5 heartbeats)²¹. Lewandowski et al. defined a four-point scale in the treatment of HCC known as the subjective angiographic chemoembolization endpoint (SACE) scale [Table 2]²². The authors reported that a sub-stasis endpoint (SACE levels 2 and 3) during TACE improved survival compared to embolization to a more complete stasis endpoint (SACE level 4)²³. Post procedural CT is helpful in determining lipiodol retention patterns, another important treatment endpoint, and a potential surrogate for treatment response in HCC²⁴. On the other hand, incomplete embolization of HCC increases plasma levels of angiogenic factors; vascular endothelial growth factor receptor (VEGFR) and insulin-like growth factor receptor 2 (IGFR-2), which have been associated with the development of metastasis after embolization²⁵. If near stasis is not achieved by the end of TACE, embolization can be continued with bland particles until the desired endpoint is achieved, the procedure can be combined with ablation, or a repeat procedure can be scheduled.

Table 2.

Subjective angiographic chemoembolization endpoint (SACE) scale²²

Grade	Antegrade arterial flow	Tumor blush
I	Normal	Normal / Reduced
II	Reduced	Reduced
III	Reduced	None
IV	None	None

Technical endpoints for liver metastases are more variable, with some texts advising embolization to stasis for bland embolization, while others recommend embolization to sub-stasis in TACE^12,26. Following embolotherapy, cross-sectional imaging is usually performed one-month post procedure, though this practice varies by operator and center. Tumor response is typically radiologically assessed by the modified Response Evaluation Criteria in Solid Tumors (mRECIST) or European Association for the study the Liver (EASL) criteria²⁷.

Recently, there has been increased interest in embolotherapy using flow / pressure modulation in parallel with the development of anti-reflux catheters. New anti-reflux catheter devices may improve therapeutic response from both safety and efficacy perspective. Two strategies have been developed to improve delivery of embolic agents to the tumor; balloon-occlusion (B-TACE) and micro-valve catheters (MVI-TACE). Balloon occlusion of the artery supplying the tumor preferentially re-routes blood flow to the tumor. These hemodynamic alterations allow increased embolic agent delivery and is based on fluid mechanic principles whereby blood flows from high to low pressure^28–30. It has been shown usage of balloon-occlusion catheters in the treatment of HCC can result in much denser lipiodol uptake in tumors, correlating with treatment response^31,32.

A micro-valve catheter has an expandable tip that collapses during forward flow and then dynamically seals off the vessel with reversal of flow, analogous to a valve. The expandable tip will collapse except if the pressure in the tip is greater than systolic blood pressure, in which case the tip will expand and continue to seal the vessel³³. TACE using micro-valve catheters has demonstrated a higher pathologic HCC tumor response compared to standard end-hole catheters³⁴.

Patient Selection

Evaluation of patients with liver cancer includes a detailed history and physical exam, as well as laboratory and imaging tests to assess generalized performance status, hepatic function, tumor burden, and vascular anatomy. Patients with an incompetent or recently cannulated Sphincter of Oddi are often given pre-procedure and post-procedure antibiotics. While controversial, many standard contraindications are considered relative depending on the operator, degree of planned treatment, presence and extent of comorbidities, and overall clinical scenario. While dependent on therapeutic modality and operator, general contraindications to embolotherapy are detailed in Table 3^35–37.

Table 3.

General Contraindications to Hepatic Embolotherapy

Patient factors

Performance status (Eastern Cooperative Oncology Group > 2) Renal failure (creatinine ≥2.0 mg/dL) Unfavorable anatomy (inability to prevent arteriovenous shunting or non-target embolization)

Liver factors

Decompensated liver disease (Child Pugh B ≥ 8) Alterations to portal vein flow (transjugular intrahepatic portosystemic shunt, thrombosis, hepatofugal flow)

Tumor factors

Extensive tumor with complete replacement of both lobes (> 75%)

Applications of Embolotherapy by Liver Tumor Type

Primary Liver Cancer

Hepatocellular carcinoma (HCC)

HCC is the most common primary liver tumor, accounting for eighty percent¹. Its incidence and prevalence are increasing worldwide, likely as a result of increasing alcohol use and obesity rates, in spite of improvements in treatments for infectious hepatitis³⁸. The staging of HCC is multifaceted and though multiple staging systems exist, the Barcelona Clinic Liver Cancer (BCLC) Staging and Treatment Strategy is one of the most common and widely accepted staging paradigms³⁹. Within the BCLC system, HCC amenable to treatment an be categorized as very early stage, early stage, intermediate stage, and advanced. Embolotherapy serves an important role in the management of HCC across all stages of disease.

Very Early Stage (0) or Early Stage (A)

In early and very early stage patients, where curative surgery or transplantation is not an immediate option, embolotherapy can serve the following roles: increase future liver reserve, provide a bridge therapy to transplant, or serve as a curative treatment in unsuitable surgical candidates.

Increase future liver remnant (FLR)

Embolotherapy—either alone or in combination with portal vein embolization (PVE)—can induce future liver remnant (FLR) hypertrophy while simultaneously providing local disease control^40,41. A systematic review and meta-analysis demonstrated an increased surgical resection rate in sequential cTACE and PVE than PVE alone for HCC > 5 cm (90 % vs. 75 %; P < 0.001)⁴¹. Furthermore, combined cTACE and PVE was associated with better overall survival and recurrence-free survival than PVE alone for patients waiting to undergo major hepatectomy,^42,43.

Radiation lobectomy, can also induce FLR hypertrophy and provide local tumor control with volumetric changes comparable to portal vein embolization, albeit slightly slower⁴⁴. A systematic review, constituting a sub-cohort of 215 patients with HCC (out of 312 patients), found rates of contralateral liver lobe hypertrophy following radiation lobectomy ranging from 26 – 47% at a range of time to hypertrophy from 44 days to 9 months. One of the included studies compared PVE to radiation lobectomy directly, describing significantly greater hypertrophy in the PVE group—61.5% versus 29.0% (P<0.001)—with a shorter median time to hypertrophy of 33 days (range 24–56 days) versus 46 days (range 27–79 days)⁴⁵. Resection following radiation lobectomy has been found to be safe and feasible, offering “curative” surgery to patients initially thought to have unresectable tumors⁴⁶.

Bridge to transplant

Embolotherapy can decrease serum alpha-fetoprotein (AFP), reduce organ waitlist drop-off due to tumor progression or death, as well as downstage tumors to transplantable disease. Bridging therapy is estimated to reduce dropout rate to 0 – 10% from up to 19.5% at 12 months^47,48. All embolotherapies have shown similar safety and efficacy profiles for patients requiring bridging therapy^49–51. Although there was no difference in overall survival, a prospective study of patients with either BCLC Stage A or B comparing SIRT to cTACE observed longer times to progression (TTP) in the SIRT group in comparison to the cTACE group – >26 months in the SIRT group versus 6.8 months in the cTACE group, P = 0.012, (HR 0.122, 95% CI, 0.027–0.557, P=.007). The authors concluded that the longer TTP provided by radioembolization could potentially reduce dropout from transplant waitlists. However, there was no significant difference in tumor necrosis and median survival times⁵².

Pathologic response to locoregional techniques may also have important prognostic consequences: complete pathologic response (CPN) of the treated tumor on explant is associated with reduced HCC recurrence and improved post-transplant survival^53,54. The use of specialized catheter delivery tools such as balloon and microvalve catheters may further augment tumor response for appropriate candidates in this setting. CTACE performed with micro-valve catheters prior to transplant for HCC demonstrated significantly higher percentage tumor necrosis on explant, 89.0±2.2%, compared to standard end-hole catheters, 56.1±44.5% (P=0.006)³⁴. Locoregional therapy performed prior to transplantation can provide useful prognostic information. Candidates with an ineffective response or interval progression can be removed from the waiting list based on anticipated long term prognosis⁵⁵.

Alternative to surgery / ablation

Ablation is recommended for patients with unresectable tumor(s) or contraindications to surgical resection or transplantation for very early and early stage HCC^56,57. In cases where ablation is not feasible, embolotherapy offers an alternative locoregional option. A retrospective study achieved similar 5-year overall survival for patients with a single-nodule HCC of 3 cm or smaller, without vascular invasion, if treated by resection, ablation or TACE⁵⁸.

In cases where TACE is performed as first line therapy, initial objective response (complete response and partial response by mRECIST criteria) after the first session is the most robust predictor of favorable outcome²⁷. In solitary tumors ≤5 cm, an ultraselective approach is preferred. Ultraselective cTACE is defined as cTACE administered at the most distal portion of the subsubsegmental hepatic artery⁵⁹. This is because well-differentiated tumors or tumors with (extra)capsular and microsatellite tumors are also supplied by the small arterial collaterals and the portal vein⁶⁰. If not completely treated, recurrent or residual tumor may transform to a more aggressive grade^61–63. Ultraselective cTACE facilitates maximal lipiodol injection into the tumor sinusoids by physiological blood flow and mechanical blockage of backflow by the microcatheter, which then pervades into the portal system⁵⁹. Then, subsequent Gelfoam® embolization occludes the hepatic artery and additional arterial collaterals, and prevents washout of the chemoembolic emulsion. Given higher rates of CPN using B-TACE or MVI-TACE as discussed above, efficacy of TACE in such scenarios can potentially be improved for appropriate candidates. The precise mechanism is not known, though may be secondary to flow diversion and delivery via small intra-hepatic collaterals which would not be apparent using standard end-hole techniques, or delivery of higher doses of therapeutic agent deeper into the tumor by overcoming intra-tumoral pressure barriers^34,64. Dual occlusion balloon catheter systems may also facilitate embolotherapy to hypovascular or infiltrative tumors that have tiny feeding vessels that cannot be selectively catheterized, such as those arising from a main arterial pedicle, otherwise precluding both non-selective and selective delivery²⁷.

Radiation segmentectomy may be employed as an alternative to percutaneous ablation, particularly in patients with unfavorable anatomic location of the target lesion. Lewandowski et al. performed radiation segmentectomy in 70 patients with solitary HCC, less than or equal to 5 cm, not amenable to surgery or ablation and Child-Pugh class A liver function⁶⁵. 59% of tumors demonstrated a complete response by European Association for the Study of the Liver (EASL). The median TTP recorded was 2.4 years (95% CI: 2.1 – 5.7 years) and median overall survival was 6.7 years (95% CI: 3.1 – 6.7 years). The authors determined radiation segmentectomy provided response rates and survival outcomes comparable to curative-intent treatments for selected patients with early-stage HCC.

Intermediate Stage (B)

BCLC intermediate stage accounts for 20 % of HCC patients with a 2-year survival of approximately 50 %⁶⁶. It is important to note the extensive heterogeneity of this patient sub-cohort given the potential range of tumor numbers and sizes as well as range of hepatic function. Patients in the intermediate stage are stratified according to these parameters, with therapeutic indication either as a means of downstaging to potential surgery or as palliative therapy to delay progression and prolong overall survival.

Downstage to Transplant

The Milan criteria is stringent in order to identify the optimal candidates for transplant with the lowest risk of tumor recurrence. However, many patients are ineligible for transplantation due to tumor burden that extends beyond these size and number criteria. Therefore, locoregional therapy is often attempted to downstage patients to transplant eligibility, and to provide a test of tumor biology prior to work up for transplant. Historically, TACE has been the most widely used downstaging therapy, however, SIRT or TAE also provides a viable alternative. A systematic review and meta-analysis was performed to evaluate the success of downstaging to Milan criteria, post-transplant recurrence and overall survival⁶⁷. In patients with HCC without vascular invasion, a pooled TACE and SIRT success rate of 54% (95% CI: 45% – 63%) was reported, with no difference between rates of successful downstaging between (P = 0.51). The pooled post-transplant recurrence rate was 0.16 (95% CI: 0.11 – 0.23), with no difference between modalities (P = 0.40). However, there was moderate heterogeneity between the included studies, and results were further hampered by inconsistency in patient selection, technique, and outcome assessments. The authors also noted heterogeneous and inconsistently reported wait-times and substantial variations in the time-to-recurrence. Post-transplant recurrence rates and survival data are difficult to objectively examine due to heterogenous patient populations and administered locoregional therapies⁶⁸. Multicenter, well-controlled studies are recommended though logistically challenging⁶⁸.

Palliative treatment of multinodular HCC

cTACE is the historically recommended therapy for patients with intermediate stage HCC in the BCLC system. This recommendation stems from seminal work by Llovet et al. which demonstrated increased survival in patients with unresectable HCC, as well as preserved liver function (CP score < 9) and performance status (ECOG 0), treated by cTACE compared to conservative management. Survival probabilities at 1 year and 2 years were 82% and 63% for cTACE, and 63% and 27% for control (P = 0.009). The SPACE clinical trial attempted to build on this by investigating potential improvement in efficacy and safety of DEB-TACE when combined with sorafenib by comparing to DEB-TACE alone in intermediate stage HCC⁶⁹. The combined arm did not delay time to progression or prolong overall survival in patients, while appearing to increase expected adverse events.

There is increasing use of SIRT for unresectable HCC. A meta-analysis of three randomized controlled trials, including the SIRTACE and PREMIERE trials, found no difference in terms of progression free or overall survival between TACE and SIRT. Of note, however, the PREMIERE trial was one of the first randomized controlled trials to compare SIRT to TACE in BCLC Stage A/B patients, and demonstrated a longer time to progression, 14.5 months (SIRT) versus 6.4 months (TACE) (P = 0.0019) without significant difference in overall survival (23.8 months vs 17.7 months, P = 0.9772)⁷⁰. On evaluation of the three trials, there were significant study limitations, such as unmet predetermined sample size and inconsistent results, probably due to differences in study populations, BCLC stage, and reported outcomes⁷¹. However, in a study examining patients refractory to TACE, 36.7 % patients underwent SIRT and demonstrated a response on imaging, with 10% successfully downstaged received liver transplantation⁷².

Advanced Stage (C)

Advanced stage HCC carries a dismal prognosis with a median survival of 2 – 4 months [61]. Systemic treatment is the recommended for patients with portal vein invasion⁷³. Sorafenib confers a modest survival increase at the expense of considerable drug-related toxicity, most commonly, diarrhea, weight loss, hand–foot skin reaction, and hypophosphatemia⁷⁴.

Historically, embolotherapy had been contraindicated in portal vein thrombosis, as TACE administered as lobar and even whole-liver treatments carried the theoretical risk of ischemia, infarction and worsening of liver function⁷⁵. A systematic review and meta-analysis examined the efficacy of cTACE stratified by main (MPV) and branch (PVB) portal vein invasion⁷⁶. 1333 patients received an estimated median of 2 (Range 2 – 4) TACE sessions. Response rates, defined by complete or partial response according to the mRECIST criteria, were similar; 14% (9 – 20%) and 16% (8% – 26%) (P = 0.238). However, results demonstrated a 1-year survival rate of 16 % (95% CI: 7% – 28%) and 37% (23 – 51%), for MPV and PVB, respectively (P < 0.001). Pooled liver failure and complication (not otherwise specified) rates were 1% and 18%, respectively. The authors attribute the improved survival of cTACE to improvements in microcatheter technology, facilitating superselective and ultraselective cTACE. By delivering cTACE in such a fashion, arterial blood flow to normal hepatic parenchyma can be preserved, minimizing the risk of the infarction of normal hepatic parenchyma and subsequent liver failure. Additionally, migration of the lipophilic chemoembolic emulsion into the portal system may theoretically provide additional antitumor effects.

It was postulated that administration of sorafenib in combination with TACE might be useful to counter act TACE-induced angiogenic factors and, therefore, improve outcomes from TACE treatment⁷⁷. This appeared to be confirmed in a systematic review and meta-analysis, demonstrating TACE-sorafenib combination therapy significantly improved time to progression (HR=0.66, 95% CI 0.50–0.81, P=0.002) with a trend to improved overall survival (HR=0.63, 95% CI 0.55–0.71, P=0.058), compared to TACE alone. Also, the disease control rate, defined as a combination of complete response rate, partial response rate and stable disease rate, increased significantly in the combination therapy group (OR=2.93, 95% CI 1.59–5.41, P=0.005)⁷⁸.

The role of SIRT in advanced HCC remains undetermined in BCLC guidelines⁷⁹. The SARAH trial failed to show superiority of SIRT to sorafenib, in 467 patients with BCLC Stage A – C, recurrent tumor ineligible for curative surgical therapy or tumors refractory to transarterial chemoembolization, Child-Pugh score A – B and ECOG 0 – 1⁸⁰. Similarly, the SORAMIC trial failed to illustrate a survival benefit of SIRT-sorafenib combined therapy in BCLC B – C patients not eligible for TACE⁸¹.

When comparing SIRT to sorafenib in patients with HCC and portal invasion, a study examining 151 patients with HCC and portal vein tumor thrombus found an overall median survival benefit in patients treated with SIRT compared to sorafenib, 26.2 months versus 8.7 months, respectively (HR 0.40 95 % CI: 0.19 – 0.82) (P = 0.013). The difference in overall survival was more pronounced in patients with branch portal vein thrombosis (as opposed to main portal vein thrombosis), with a respective median OS of 25.3 months (95 % CI 13.8– 36.8) vs 7.0 months (95 % CI 5.2–8.9) (p=0.001), than in cases of main PVT, with a respective median OS of 12.0 months (95 % CI 4.6–19.3) vs 6.5 months (95 % CI 4.8–8.3) (p=0.195)⁸². SIRT has also been shown to preserve health-related quality of life in patients with HCC and portal vein invasion in comparison to systemic therapy⁸³.

Intrahepatic cholangiocarcinoma (ICC)

ICC accounts for 10 – 20% of primary liver tumors, though the incidence is rising⁸⁴. Surgery is the only curative therapy; however, unlike perihilar or distal bile duct tumors, intrahepatic lesions are often asymptomatic and often present with locally advanced tumours beyond the threshold of surgical resection⁸⁵. Overall survival is 11.7 months with systemic chemotherapy⁸⁶. The National Comprehensive Cancer Network (NCCN) guidelines recommend consideration of locoregional therapies in cases of unresectable or metastatic ICC²⁵.

A systematic review examined hepatic embolotherapy, specifically TACE and SIRT, in 929 patients with unresectable ICC⁸⁷. The majority of the included studies were retrospective with small patient cohorts, highlighting the need for prospective trials in this patient subset. There were variable prior chemotherapy and surgical treatments in the different studies. TACE regimens lacked standardization of technique, from chemotherapeutic agents delivered to number of sessions. Median duration of follow up was 35 months (Ranged 1.8 – 29 months). Complete, partial response, and stable disease rates for TACE were 10%, 20% and 57%, respectively, and for SIRT were 10, 25.5% and 66.5%, respectively. Progression free and overall survival for TACE were 8 months and 13 months, respectively, and for SIRT were 9.8 months and 12.5 months, respectively.

Zhen et al. performed a pooled analysis of 472 patients who received SIRT for unresectable intrahepatic cholangiocarcinoma⁸⁸. The majority of the included studies were retrospective with small patient cohorts. The overall response rate was 11.5% (Range: 4.8 – 35.3%) The pooled median overall survival was 14.3 months (95% CI: 11.9 – 17.1 months). OS appeared consistent across studies without significant study heterogeneity, I² = 50.9% (P= 0.068). However, not all studies reported the population and treatment characteristics. There was no difference in OS between glass or resin microspheres. Most common side effects were fatigue (median: 31.7%; range: 0.0–87.5%), abdominal pain (median: 30.0%; range: 0.0–85.0%) and nausea (median: 16.0%; range: 0.0–62.5%). Grade III – IV toxicities were reported in 10 studies, median of 7.8% (range: 0.0–25.0%). 10% of patients in three selected studies were converted to resectable disease based on a systematic review and pooled analysis of unresectable ICC, and is therefore now presented as an option to patients with advanced disease in the latest European Society for Medical Oncology (ESMO) guidelines⁸⁹. TACE and SIRT are safe and efficacious in the treatment of unresectable ICC, but, their exact role in a therapeutic paradigm remains to be defined.

Colorectal liver metastases

Colorectal cancer is the third most common cancer in the United States⁹⁰, with 15% of presenting with synchronous hepatic metastases, and 50% of patients with subsequent metastases⁹¹. Hepatic metastectomy in the setting of colorectal liver metastasis (CRLM) has been established as a standard of care through comprehensive, retrospective, large population-based studies incorporating surgery and/or ablation demonstrating 5-year survival of 30–50% in patients undergoing surgical resection as compared with 0% 5-year survival in patients that forgo surgery^92,93. Yet, only 20 – 30% of patients will have resectable disease, of which 70–80% will have recurrence within 5 years^3–5. For unresectable CRLM, systemic therapy has demonstrated overall survival of 15–20 months, with mortality predominantly attributed to hepatic progression^94–96.

The NCCN guidelines recommend hepatic resection for resectable liver metastases from colorectal cancer⁹⁷. Currently, in patients with liver-only / liver-dominant metastases that cannot be surgically resected or ablated, transarterial therapies may be considered, particularly in patients who have progressed on first line therapy. The ESMO guidelines state radioembolization should be considered and chemoembolization may be considered in patients with liver-confined disease refractory to systemic chemotherapy⁹⁸. However, there may be a role for embolotherapy in the neoadjuvant and adjuvant settings⁹⁹.

Two randomized controlled trials have been performed, comparing DEB-IRI TACE to different systemic chemotherapy regimens^96,100. Martin et al. conducted a phase II clinical trial assessing the efficacy and safety of DEBIRI in combination with systemic modified FOLFOX (folate, fluorouracil and oxaliplatin) ± bevacizumab as compared to FOLFOX ± bevacizumab alone in chemotherapy naïve patients. Seventy patients (40 in the FOLFOXDEBIRI arm and 30 in the control arm) were included in the study. For each DEBIRI treatment, 100 mg of irinotecan-loaded 100–300 μm LC Beads® (BTG, London, UK) were delivered to each lobe. Patients received at least 12 cycles of systemic treatment and two cycles of DEBIRI during chemotherapy off weeks. The median follow-up was 24 months. The overall response rate (CR or PR by mRECIST) was greater in the treatment arm versus the control arm at 2 (78% vs 54%, P=0.02), 4 (95% vs 70%, P=0.03), and 6 months (76% vs 60%, P=0.05). This favorable overall response rate in the treatment arm at 6 months led to more downstaging to resection, 35% versus 16% (P=0.05) and longer median progression-free survival, 15.3 vs 7.6 months.

There was a nonsignificant improvement in median progression-free survival in the control arm, 15 months (Range 10.4 – 20 months) versus the treatment arm, 12 months (Range 9 – 15.4 months) (P=0.18). It should be noted that there were significant baseline differences in disease severity between the study arms in favor of the treatment arm: there was a statistically significantly worse overall Eastern Cooperative Oncology Group performance status, prevalence of extrahepatic disease, and prevalence of KRAS mutant disease in the control arm.

There were more serious adverse events in the FOLFOXDEBIRI arm relative to the FOLFOX alone arm. These were attributed to embolization-related serious adverse events, specifically defined as hospital stay > 23 hours or readmission after DEBIRI treatment, which accounted for 33% of such events. When evaluated only for chemotherapy-associated adverse events, there was no difference between the study arms.

Fiorentini et al. reported their findings in a prospective randomized controlled trial comparing DEBIRI to irinotecan, fluorouracil and leucovorin (FOLFIRI) in patients whom had previously received chemotherapy, but not irinotecan. Seventy-four patients with liver-only metastatic disease were enrolled in the study (36 in the DEBIRI arm and 38 in the FOLFIRI arms), all of whom had received 2–3 lines of prior chemotherapy. The study aimed to deliver 200 mg of irinotecan with 100–300 μm LC beads® two times with one month interval between treatments for the DEBIRI cohort. The FOLFIRI arm received a total of 8 cycles. A significant increase in OS was observed in the DEBIRI group, 22 months, compared to the FOLFIRI group, 15 months (P = 0.031). PFS was also longer in the DEBIRI group, median seven months, versus the FOLFIRI group, 4 months, (P = 0.006). Significantly less neutropenia and mucositis were recorded in the DEBIRI arm compared to FOLFIRI arm, 4% versus 44% (P < 0.0001), and 1% versus 20% (P = 0.00002), respectively.

Selective internal radiation therapy

A phase III clinical study demonstrated a role for SIRT as salvage therapy for patients with chemotherapy-refractory liver-limited metastatic colorectal cancer¹⁰¹. A recent combined analysis of three randomized controlled trials, involving 549 patients, investigated systemic oxaliplatin-based chemotherapy (FOLFOX: leucovorin, fluorouracil, and oxaliplatin) chemotherapy upfront with SIRT or FOLFLOX alone¹⁰². There was no difference in overall survival (HR: 1·04, 95% CI 0·90–1·19; P = 0·61). The median survival time in the FOLFOX-SIRT arm was 22.6 months (95% CI 2.0 – 24.5), versus 23.3 months (21.8 – 24.7) in the control arm. Front line use of SIRT with chemotherapy was not recommended and the authors stated the role of SIRT as consolidation therapy after chemotherapy with careful patient selection requires investigation.

In a systematic review and meta-analysis of transarterial therapies in 2306 patients, 1038 TACE and 1268 SIRT, with unresectable CLRM, first line TACE demonstrated a non-statistically significant decreased median OS of 15.2 months compared to SIRT, 29.4 months (P = 0.69)¹⁰³. For patients failing at ≥ 1 line of systemic therapy; the survival outcomes were 21.3 months (Range 20.6 – 22.4 months) in the TACE group, compared to 10.7 months (Range 9.5 – 12.0 months) in the SIRT group. Chemotherapy regimens were varied amongst the different studies, with DEBIRI accounting for only 53% of studies. TACE had higher rates of conversion to surgically resectable disease, 4% (95% CI: 2 – 7%), compared to SIRT 2% (1–4%). Combined grade 3 – 4 toxicities were seen in 26% (22–30%) of patients who received SIRT, and 17% (13–22%) of patients who underwent TACE. Levy et al conducted a more recent systematic review and meta-analysis of 1583 patients—968 TACE and 615 SIRT—with a different methodology¹⁰⁴. Results for TACE were stratified into DEBIRI and cTACE. One-year survival rates were 80.1% (95% CI: 74.2 – 86.0%) for DEBIRI, 69.6% (95% CI: 49.3 – 86.7%) for cTACE and 41% (95% CI: 28 – 54%) for SIRT. Higher 1-year survival rates may be explained by increased radiologic response (CR and PR by RECIST criteria) rates for DEBIRI, 36.2% (95% CI: 0 – 72.5%), compared to cTACE, 23.1% (95% CI: 9.7 – 36.4%). However, pooled median survival were similar 16 months (9.0–23) and 16 months (7.3–25) for DEBIRI and TACE, respectively.

Gastro-entero-pancreatic neuroendocrine tumor (GEP-NET) liver metastases

The prevalence of liver metastases from GEP-NETs vary depending on the site of the tumor. For example, metastases were most common from the pancreas (40.3%) and lung (32.4%), and least common from the rectum (3.4%) and stomach (9.4%)¹⁰⁵. The origin, grade (G1-G3) and stage of the tumor are the most important prognostic factors, with presence of metastasis having the highest risk of mortality. Patients with favourable (moderate / well differentiated) pathology and disease burden (regional / localized) can live more than 10 years. For example, only 7.3% of patients with pancreatic NETs were alive at 10 years, compared with 20.5% of those with colon NETs and 27.1% of those with rectal NETs. Treatment strategies are chosen based on three different metastatic patterns: Simple (metastases confined to one liver lobe or limited to two adjacent segments that can be resected by a standard anatomical resection), complex (metastases predominantly in one lobe primarily but with smaller satellites in the contralateral lobe) and diffuse (multifocal metastases). Surgery and/or ablation is generally indicated for simple and complex patterns, with embolotherapy traditionally reserved for diffuse pattern metastases¹⁰⁶.

Embolotherapy in the setting of GEP-NET liver metastases is indicated for the control of tumor growth or control of secretory symptoms after the failure / insufficiency of somatostatin analogs, and when hepatic resection is contraindicated^106,107. Embolotherapy is recommended early in the disease course if there is significant liver tumor burden or the secretory symptoms are life-threatening or severe in intensity. The European Neuroendcrine Tumor Society (ENETS) recommend embolotherapy only for G1 – G2 tumors, while systemic chemotherapy is typically reserved for neuroendocrine carcinomas (G3).

GEP-NETs are more arterialized than other liver metastases, such as from colorectal cancer¹⁰⁸. No specific embolotherapeutic modality is recommended over the other by the NCCN or ENETS, though the choice of embolotherapy should be balanced between effective tumor control and minimizing damage to normal liver. Symptomatic control ranges from 42% – 91% in TAE and 54 – 100% in TACE¹⁰⁹. Progression free survival and overall survival ranges from 6 – 36 months and 16 – 69 months in TAE and 5 – 32 months and 15 – 65 months in TACE, respectively. A small randomized trial of 26 patients with progressive unresectable liver metastases from midgut endocrine tumors were randomly assigned bland embolization or chemoembolization and examined for differences in 2-year PFS rate¹¹⁰. A subgroup analysis of one study with 69 patients with G1 NETs and 54 patients with functional islet cell carcinomas showed the former had longer PFS, 22.7 months versus 16.1 months (P = 0.046), and OS, 33.8 months vs. 23.2 months (P = 0.012)¹¹¹. The authors observed a possible beneficial outcome by delivering chemoembolization in patients with islet cell carcinomas. DEB-TACE is specifically known to cause biliary / liver injury. In a series of 208 patients with either well-differentiated metastatic NET or HCC, treated by cTACE or DEB-TACE, biloma / parenchymal infarct was strongly associated with both DEB-TACE (OR = 9.78, P = 0.002) and NETs (OR = 8.13, P = 0.04)¹¹².

Analogous to HCC, increased levels of angiogenic factors and neoangiogenesis from embolization can drive growth of incompletely treated tumors^113,114. A retrospective review examined whether using different size bland microspheres in 160 patients with well to moderately differentiated NET liver metastases¹¹⁵. The hypothesis was whether increased occlusion of end arterioles and subsequent ischemia would lead to improved overall survival. Higher objective response rates in the smaller particle group (< 100 μm) compared to the larger particle group (≥100 μm), 64 versus 42% (P=0.007), respectively, did not translate in to prolonged overall survival. A phase II clinical study administering sunitinib, a tyrosine kinase inhibitor, following embolotherapy in 39 patients with well/moderately differentiated NETs from a range of primary tumor origins (26 patients had small bowel primary tumor) yielded a median PFS of 15.2 months and 1-year and 4-year survival rates of 95% and 59%, respectively¹¹⁶. Another phase II clinical study examined whether everolimus, an mTOR pathway inhibitor, could increase hepatic-PFS at 24 months from 35 to 50 %, when administered after embolotherapy¹¹³. The primary endpoint was not reached.

SIRT has can also be employed for the treatment of neuroendocrine tumors, though may potentially manifest undesired sequelae in patients with longer prognoses. The most common early adverse events reported following SIRT are fatigue, abdominal pain, fever and nausea¹⁰⁹. Su et al. followed 54 patients treated with SIRT (15 unilobar and 39 whole-liver) for a mean of 4.1 years¹¹⁷. The authors observed imaging cirrhosis-like morphology in 26.7% and 56.4% unilobar and bilobar treatments, respectively. Features of portal hypertension; such as ascites (33% patients) and varices (31% patients), were seen with both more common in whole-liver treatments. Eight patients (20.5%) treated with whole-liver SIRT and exhibited cirrhosis-like morphology showed clinical signs of hepatic decompensation. Tomozawa et al. followed 52 patients for > 1 year treated with SIRT for hepatic metastases from NETs and discovered 15 patients with cirrhosis-like morphology on imaging¹¹⁸. Similar to the study by Su et al., this was more common in bilobar treatments, with significantly elevated liver function tests and thrombocytopenia in this sub-cohort, sustained >4 years after treatment. No significant difference was found in predicting overall survival between unilobar and bilobar treatments (P = 0.676).

A review of the current literature reports successful secretory symptom control rates with embolotherapy ranging from 55% – 100%¹⁰⁹. The progression-free survival and the overall survival ranges are 9 – 11 months and 14 – 70 months, respectively¹⁰⁹. Factors associated with increased response rate and / or prolonged progression free survival include female gender, radiological hypervascular pattern, well-differentiated tumors and a low Ki67 index¹¹⁹. A small scale prospective study comparing treatment of metastases from GEP-NETs to CRLM by SIRT demonstrated higher objective response rates in the GEP-NET group, 73% versus 25%, and prolonged progression free survival, 1.0 months versus 8.9 months, and overall survival, 6.7 months versus 4.4 months. However, no formal randomized trial has been conducted comparing the efficacy and safety of these therapies.

Other Hepatic Metastases

Liver directed therapy in patients with metastatic disease from non-GI primaries is more controversial, and the evidence more limited. The evidence is perhaps strongest for metastatic breast cancer and melanoma, though treatment has been reported for other primaries such as RCC and lung cancer^120–123. Approximately 50% of all women with breast cancer develop metastases, of which approximately 50% of patients develop in the liver. Additionally, 5–12% of patients with recurrence develop liver metastases as the primary site recurrence. Generally, embolotherapy is offered to patients with metastatic disease refractory to multiple lines of chemotherapy. A systematic review and pooled analysis of 519 patients with liver metastases from breast cancer treated by TACE (cTACE and DEBIRI) reported median overall survival ranged from 7.3 – 47.0 months and median disease free survival ranged from 2.9 to 17.0 months¹²². Several retrospective studies have investigated the efficacy of SIRT in patients with liver metastases refractory to systemic treatment, with a median OS of 4 – 13.6 months¹²⁴.

Uveal melanoma has a distinct propensity to metastasize to the liver, with up to 50% of patients developing hepatic metastases¹²⁵. Median overall survival of patients with hepatic metastases is reportedly 4 – 5 months, with a 1-year survival of 10%–15%¹²⁶. Embolotherapies are employed due to the lack of effective systemic chemotherapy¹². A review of 649 patients who received TACE, reported that the overall survival ranged from 5 – 29 months¹²³. Chemotherapy regimens varied amongst a combination of retrospective and prospective studies. TACE appeared to be a relatively safe procedure, with eight patients developing grade 3 or 4 complications, three of which were vascular thromboses that precluded further transarterial therapy. Six retrospective studies of 405 patients whom received SIRT reported the median OS ranged from 9 – 24 months¹²³. Currently, a phase I clinical trial evaluating SIRT with sorafenib [NCT01893099] and an early phase I study investigating SIRT with ipilimumab [NCT01730157] are underway.

Future therapies

Precision medicine is a significant innovation in cancer care. One specific area of interest lies in the role of immunotherapy and potential synergy with locoregional therapy. Immune cells and T-cell infiltration correlate with HCC outcome and prognosis, respectively¹²⁷. Nivolumab and Pembrolizumab, two immune checkpoint inhibitors, have been approved for second line systemic therapy in patients with advanced HCC after failing sorafenib⁵⁷. Trials have shown that the immune response in HCC elicited after treatment by ablation/embolization can be enhanced by PD-1 inhibition¹²⁸. A number of trials are being conducted to evaluate potential synergistic therapeutic effects between anti-PD1 inhibitors and transarterial therapy such as SIRT and TACE¹²⁹. Preliminary studies assessing safety of SIRT-PD1 inhibitor combination therapy has been shown to be safe¹³⁰.

Another form of precision medicine known as theranostics involves the use of specialized diagnostic and therapeutic agents such as nanoparticles or radiopharmaceuticals to diagnose and treat tumors¹³¹. The detection of particular targets in a tumor can help predict whether a patient will benefit from a targeted treatment. Typical uses of theranostics today in hepatic oncology include 177Lu-DOTATATE for metastatic GEP-NETs¹³². However, the success of theranostic agents are in large part dependent on targeting selectivity and efficiency when delivered systemically while minimizing non-target toxicity, particularly in the kidney, a potential dose limiting organ^133,134. Catheter based delivery of theranostic agents may be a way to obviate such limitations. This strategy has the advantage of delivering high doses of nanoparticles locoregionally and minimizing systemic toxicity. In a retrospective analysis of 55 patients with GEP-NETs, intra-arterial injection of peptide receptor radionuclide therapy demonstrated a median PFS and OS were 33.4 and 75.8 months, respectively, in patients with liver-confined tumors¹³⁵. Real-time treatment response monitoring can also be performed using theranostic agents and positron emission tomography/computed tomography (PET-CT) and magnetic resonance imaging depending on the specific imaging properties^136,137.

Artificial intelligence has the potential to revolutionize all aspects of healthcare, particularly image-guided therapy. For example, incomplete HCC treatment via TACE can result in increased levels of angiogenic factors, leading to tumor growth and metastasis¹³⁸. A novel study conducted by the MD Anderson Cancer Center Group demonstrated machine learning may play a role in predicting patient response to TACE using quantitative imaging features in combination with the BCLC staging system¹³⁹. Other machine learning applications in interventional oncology include intra-procedural support, such as catheter guidance, and post procedural assessment, such as post-treatment MRI-based texture analysis of tumors^140,141.

Conclusion

Embolotherapy is an important tool in the management of liver tumors. The technical advancements and indications for embolotherapy have greatly expanded, providing options to patients ranging from curative treatment in some cancers to prolonging time to progression and overall survival, as well as improved quality of life over systemic therapy. Nonetheless, the quality of evidence and grade of recommendations are limited, which hinder inclusion in to mainstream clinical societal guidelines. Future trials should examine specific and clinically relevant outcomes with strict patient selection criteria. Additionally, parallel advances in immunotherapy and artificial intelligence offer the promise of new multidisciplinary treatment paradigms in the treatment of patients with liver tumors.