Role of Transarterial Radioembolization in Management of HCC

Abstract

Transarterial Radioembolization (TARE) is an emerging therapy for the management of hepatocellular carcinoma (HCC). Radioactive Y90 can be loaded onto glass or resin microparticles for intra-arterial selective internal radiation therapy. TARE is a multistage procedure requiring a mapping procedure and a treatment procedure. The mapping procedure informs operators on vascular anatomy as well as potential arteriovenous shunting. Based on the device desired, goal of treatment, vascularity of the tumor and shunting, dosage can be computed for treatment day. Overall, TARE is generally a well-tolerated procedure with promising clinical outcomes with HCC for the purposes of curative therapy, downstaging, and palliation.

1. Introduction

HCC the most common primary liver cancer and has the fourth highest cancer-related mortality [1, 2]. The five-year survival rate of primary liver cancer is second lowest to only pancreatic cancer [3]. For early-stage disease, surgical resection and orthotopic liver transplant represent curative options but a large majority of patients are poor surgical candidates or transplant ineligible. For unresectable early-stage disease, percutaneous ablation can achieve outcomes similar to surgical resection [4]. For intermediate-stage disease, transarterial chemoembolization (TACE) offers improved survival benefits. For advanced-stage disease, systemic therapy offers modest benefits.

Transarterial radioembolization (TARE) with Yttrium 90 (Y90) is a popular and relatively new locoregional therapy (LRT) for unresectable hepatocellular carcinoma (HCC). Strategies for use include curative intent, downstaging and/or bridging to liver transplantation, and palliation. As such, TARE has role for management in HCC for unresectable early, intermediate, and advanced-stage disease. This chapter reviews TARE treatment selection amongst various LRTs in addition to treatment planning, dosimetry, safety and efficacy data, and therapeutic applications.

2. HCC and the treatment paradigm

HCC is a major world health issue, with 750,000 new cases a year and the incidence is expected to continue to rise with the rise of hepatitis C and obesity/non-alcoholic hepatic steatosis [1]. The Barcelona Clinic Liver Cancer (BCLC) system is a widely accepted staging system for HCC management, seen in Figure 1. The 2022 Update of the BCLC treatment algorithm offered an expanded role for TARE [4]. For early stage (A), TARE is used for unresectable disease in order to bridge patients for transplant. For intermediate stage disease (B), transarterial therapies including TACE and TARE are used. TARE has demonstrated longer time to progression compared to TACE (>26 months vs 6.8 months, p = 0.0012) in a randomized phase 2 study (PREMIERE) [5]. Furthermore, TARE has demonstrated a favorable adverse event profile, lower incidence of diarrhea (0% vs 21%, p = 0.031) and hypoalbuminemia (4% vs 58%, p < 0.001), and increased quality of life compared to TACE using subjective measures of social well-being, functional well-being, and embolotherapy specific toxicities [5, 6]. Lastly for advanced stage (C), TARE has shown comparable survival benefits compared to Sorafenib, a systemic chemotherapeutic, with respect to overall survival (8.0 vs 9.9 months, p = 0.18) and improved treatment tolerance with twice as many adverse events with Sorafenib in the SARAH randomized control trial [7]. The focus of the next section will be an overview of the pathophysiology and different devices used with TARE.

Figure 1.

BCLC staging system for treatment of HCC.

2.1 Pathophysiology of TARE and current devices

Transarterial radioembolization (TARE) is a form of selective internal radiation therapy described as transcatheter intra-arterial delivery of radioactive microparticles for the treatment of primary or secondary liver cancer [8]. While normal hepatic parenchyma derives >70% of its blood supply from the portal vein, primary HCC receives greater than 90% of its tumoral blood supply from the hepatic artery by way of neo-angiogenesis. This principle is utilized in TARE to safely deliver radioactive microparticles to the tumor without affecting the surrounding liver [9]. Since the PREMIERE trial in 2016, there has been further investigation and increased acceptance of TARE as a key treatment modality for HCC [5]. Yttrium-90 (Y90) is the radioactive isotope most often used in TARE. Y90 has a half-life of 64.2 h and emits pure β particles (cytotoxic to tumor cells) as it decays into Zirconium-90. The average tissue penetration of Y90 is 2.5 mm with a maximum of 11 mm. This narrow range of penetrance is important for local tissue delivery with sparing of the surrounding liver tissue [10, 11]. There are two main devices for loading of the radioactive isotope to microparticles: resin and glass microspheres which are commercially produced by SIR-spheres® (Sirtex Medical, Woburn, MA, USA) and TheraSphere® (Boston Scientific, Marlborough, MA, USA), respectively [8]. Glass microspheres are US Food and Drug Administration (FDA)-approved for the treatment of unresectable HCC or HCC complicated by portal vein invasion [12]. The porous nature of glass microspheres allows for higher loading dose of Y90 per sphere making it a suitable candidate for locally delivering the desired radiation dose without a detectable embolizing effect [8, 13]. On the other hand, resin microspheres have lower radiation dose per sphere as the radioisotope is coated on the periphery of the sphere. Therefore, a higher quantity of spheres is required to deliver a desired dose, increasing risk for stasis of blood flow and embolization of the artery before completion of instillation [9, 14]. The two devices are summarized in Table 1.

Glass microspheres	Resin microspheres
Diameter of 20–30 μm and activity of 2500 Bq per sphere [8].	Diameter of 20–60 μm and activity of 50 Bq per sphere [9].
1.2 million glass microspheres produce an activity of 3 GBq (2500 Bq per sphere) [8].	40–80 million resin microspheres produce an activity of 3 GBq (50 Bq per sphere) [8].
Indications: Unresectable HCC HCC complicated by PVT [12].	Indications: Unresectable colorectal metastases in combination with chemotherapy (fludarabine) [8].
Given the greater specific activity of Y90 per microsphere, the desired dose delivery is possible without embolizing the blood vessel [13].	Due to the lower specific activity of each microsphere, it is difficult to deliver the desired dose without embolizing the blood vessel [14].

Table 1.

Summary of Y90 devices which include glass and resin microspheres.

2.2 Treatment planning

TARE is a two-step therapy consisting of a treatment-planning procedure (“mapping”) followed by a separate intervention for Y90 administration. The mapping procedure is performed to identify the arterial supply of the treated tumor(s), assess for potential enterohepatic collaterals that may result in non-target embolization, as well as administering Technetium 99m macro aggregated albumin (Tc-99m MAA) to calculate the hepatopulmonary shunt fraction. The first stage entails hepatic angiography to evaluate the arterial anatomy. Roughly 45% of patients can have variant hepatic artery anatomy which can lead to nontarget delivery of Y90 to tissues such as the stomach, gallbladder, and bowel if not properly accounted for [15]. The common hepatic artery usually arises at the celiac trifurcation along with the left gastric artery and splenic artery. After the takeoff of the gastroduodenal artery, the common hepatic artery gives rise to the proper haptic artery which bifurcates into the right and left hepatic arteries. The two most common variations of the hepatic arterial vasculature include the replaced right hepatic artery arising from the superior mesenteric artery and the replaced left hepatic artery arising from the left gastric artery (Figure 2) [15]. Next, it is important to evaluate the origin of vessels such as the right gastric artery which typically arises from the proper hepatic artery but can arise off the left hepatic artery. If radioactive microparticles are administered proximal to the takeoff of the right gastric artery, severe radiation-induced gastric ulcerations can result [16]. Other vessels that can result in nontarget delivery of Y90 include the gastroduodenal, cystic, falciform, left inferior phrenic, and supraduodenal arteries [17]. Additionally, accessory arteries such as the accessory left gastric artery can arise from the left hepatic artery and cause nontarget embolization (Figure 2).

Figure 2.

A: Digital subtraction angiogram (DSA) of the common hepatic artery (A) with branches labeled as followed: B = proper hepatic artery; C = left hepatic artery (C-I = lat. branch of left hepatic artery; C-II = med. branch of left hepatic artery); D = right hepatic artery (D-I = ant. branch of right hepatic artery; D-II = post. branch of right hepatic artery); E = gastroduodenal artery; F = right gastric artery; G = Arterially enhancing tumor; B: DSA of replaced right hepatic artery (A) arising off of superior mesenteric artery (B); C: DSA of replaced left hepatic artery (B) arising off of the left gastric artery (A); D: DSA of accessory left gastric artery arising off of the left hepatic artery.

Nontarget vessels can be coiled off during the mapping procedure if the catheter cannot be positioned beyond the vessel origin for treatment. In addition to evaluating for causes of nontarget embolization, angiography is used to identify which arteries need to be treated to provide adequate arterial coverage. Typically, intra-procedure cross sectional imaging such as cone beam CT is performed to confirm arterial coverage as seen in Figure 3.

Figure 3.

A: DSA demonstrating arterially enhancing tumors; B: Cone Beam CT correlate from DSA assuring tumor has adequete arterial coverage.

Once arterial coverage is confirmed, Tc-99m MAA is delivered via microcatheter at the expected site of treatment. Tc-99m MAA is used to predict the distribution of Y90 on treatment day as it has similar tissue penetrance as Y90 with significantly lower activity [18]. After infusion of Tc-99m MAA, detection of gamma radiation is used to evaluate radioactive microparticle distribution with Single Photon Emission Computed Tomography (SPECT). The planar imaging (with or without accompanying CT) is compared to mapping angiography to assess if the bulk of gamma radiation counts are localized to area of local injection. At this time, nontarget delivery of MAA can be assessed by evaluation radioactivity in the stomach and bowel. In addition to accessory vasculature leading to nontarget embolization, reflux of microparticles during MAA can also contribute to extrahepatic delivery. Planar imaging can also be used to compute the hepatopulmonary (“lung shunt”) fraction by calculating the arithmetic or geometric mean of radioactivity counts with in the lung, using anterior and posterior projections, compared to the sum of counts in the liver and the lungs. Arteriovenous shunts can provide a direct pathway for radioactive microparticles to traverse from the liver to the lung leading to radiation pneumonitis. Sample planar imaging is seen in Figure 4 and the equation for assessing the lung shunt fraction is seen in equation 1. For glass microparticles there is a high association of radiation pneumonitis with doses of greater than 30 Gy per treatment or greater than 50 Gy cumulatively across all lifetime treatments [11]. For resin microparticles, a lung shunt fraction of less than 20% is recommended [11].

$$\mathrm{LSF}=\frac{\text{Mean Lung Counts}}{\text{Mean Lung Counts}+\text{Mean Liver Counts}}$$E1

Figure 4.

Sample planar SPECT images showing radiation counts in the lungs and liver.

Equation 1. Equation for calculation of lung shunt fraction.

2.3 Dosimetry of TARE

Dosimetry refers to the dose determination process for TARE and takes into account many factors such as intent of therapy (i.e. curative vs palliative), tumor burden (segmental vs lobar), arteriovenous shunting/lung shunting, and patient size. Lobar or radiation lobectomy refers to treatment of 3 or more liver segments whereas “segmentectomy” refers to two or fewer segments as identified by the Couinad liver segmentation classification system [11]. Additionally, it is important to understand the difference between dosage as well as activity. Dosage refers to the amount of energy (Joules) imparted on a specific mass of tissue (kg) and is measured in gray (Gy). Activity is defined by the amount of energy that is lost per unit time by an isotope due to radioactive decay and is measured by Becquerel (Bq).

There are two main mathematic models that are used to calculate dosage which include the Medical Internal Radiation Dose (MIRD) model and the Partition model. Each of these models are summarized in Eq. (2) [11]. The MIRD is a single compartment model that assumes there is uniform distribution of activity into the desired tissue and the inputs are the desired dosage for treatment, mass of the liver, and the lung shunt fraction [11]. Historically, for lobar treatment with glass microspheres a dosage of 120–150 Gy is used whereas segmental treatment dosing is done with target dosing >400 Gy using MIRD.

The Partition model assumes that the radioactive microparticles distribute into non-tumorous liver tissue as well as the intended tumor. The inputs of the Partition model include: the maximum dose that is desired to the healthy liver, the ratio of MAA uptake in the tumor vs the healthy liver which is determined by SPECT or SPECT CT, mass of tumor and normal liver tissue, and lastly the lung shunt fraction. MIRD model is used primarily for glass microparticles while the Partition model can be used for both microsphere types [19].

$$\mathrm{A}:D_{\mathit{\text{liver}}}\left[\mathit{Gy}\right]=50\left[\frac{J}{\mathit{Gbq}}\right]\times \frac{A_{\mathit{\text{total}}}\left[\mathit{GBq}\right]}{m_{\mathit{\text{liver}}}};D_{\mathit{\text{lung}}}\left[\mathit{Gy}\right]=50\left[\frac{J}{\mathit{Gbq}}\right]\times \frac{A_{\mathit{\text{total}}}\left[\mathit{Gbq}\right]}{m_{\mathit{\text{lung}}}\left[\mathit{kg}\right]}\times \mathit{LSF}$$

$$B:D_{\mathit{\text{liver}}}\left[\mathit{Gy}\right]=\frac{50\left[\frac{J}{\mathit{GBq}}\right]\times \left(A_{\mathit{\text{total}}}\right)\left[\mathit{GBq}\right]\times \left(1-\mathit{LSF}\right)}{m_{\mathit{\text{non}}-\mathit{\text{tumor}}}\left[\mathit{kg}\right]+\frac{T}{N}\times \left(m_{\mathit{\text{tumor}}}\right)\left[\mathit{kg}\right]}$$E2

Equation 2; A represents the MIRD model and B represents the partition model. A is activity, D is dose, m is mass, and LSF is lung shunt fraction.

Both of the above-mentioned equations can be used to compute total activity of Y90 needed for the intended dosage of therapy. When the microparticles are ordered, they are calibrated by activity and are usually in 3 GBq per vials for resin microparticles and 3, 5, 7, 10, 15, and 20 GBq vials for glass microparticles [11]. Each respective manufacturer reports a shelf life of 12 days for glass microparticles and 3 days for resin microparticles. Each particle, within the vial, has a specific activity which determines the overall vial activity. Each particle can either have a high specific activity or low specific activity. If a large tumor is being treated, more microparticles are needed and therefore lower specific activity microparticles are desired [19]. The opposite is true for smaller tumors where fewer microparticles with higher specific activity is optimal. Vials are selected based on patient specific dosage and tumor size as well as the exact date and time of treatment because of the radioactive decay of Y90. Resin microspheres are calibrated on treatment day while glass microspheres are calibrated in advance and must be administered at a specific time [20].

2.4 Treatment day

Patients typically return 1–2 weeks following the mapping procedure for Y90 administration. The microcatheter is placed into the tumor-feeding arteries that were identified on the mapping procedure. After Y90 administration, the catheters used for infusion as well as the administration apparatus are collected for radiation survey and disposed based on institutional radiation safety protocol [11]. Of note, there are no extra precautions advised to the patient and family members after TARE regarding external radiation exposure as Y90 is a pure beta emitter and the radiation is confined to the abdomen [21].

After the procedure, it is prudent to evaluate the radioactive microparticle distribution within the hepatic and extra-hepatic tissue, similar to the mapping procedure with Tc-99m MAA. SPECT-CT in addition Positron Emission Tomography (PET) can be used to visualized the radioisotope distribution. Y90 undergoes beta negative decay which produces bremsstrahlung photons that are detected in SPECT-CT and PET-CT imaging. SPECT-CT is more commonly performed for biodistribution evaluation however PET-CT is believed to provide more precise and accurate detection [11]. An example of SPECT-CT showing biodistribution within the right hepatic lobe after lobar infusion is seen in Figure 5.

Figure 5.

Increased radiotracer uptake in the right hepatic lobe corresponding to treated region.

2.5 Safety and efficacy of TARE

TARE with Y-90 is a generally safe and effective. Immediately after the procedure, patients may experience post embolization syndrome (PES) which is comprised of an array of symptoms including abdominal pain or discomfort, nausea, vomiting, and fatigue. PES is usually self-limiting and treatment is supportive [22].

Radiation-induced liver disease (RILD) is one of the most feared adverse events related to TARE and occurs in less than 2% of the cases [23]. RILD is defined as new onset anicteric ascites with an increase in alkaline phosphatase that is out of proportion to other biliary enzymes within 2 months of TARE [24]. Repeat treatment with TARE has been demonstrated to be safe and effective in patients with residual or recurrent HCC [25].

Biliary complications are rare but include radiation cholecystitis, radiation-induced cholangitis, biliary necrosis, obstructive jaundice secondary to biliary strictures, and abscess/bilomas [23]. Pancreatic complications are uncommon and mostly include acute pancreatitis [23]. Pulmonary complications include radiation pneumonitis, atelectasis and pleural effusion. Gastrointestinal adverse events include diarrhea and mucosal ulceration which are consequences of nontarget embolization from hepatoenteric arterial communications [23].

The efficacy of TARE is assessed via cross-sectional abdominal imaging such as contrast-enhanced CT or MRI. Various imaging criteria exist to evaluate treatment response, including the modified Response Evaluation Criteria in Solid Tumors (mRECIST) and Liver Imaging Reporting and Data System (LI-RADS) [26, 27]. There is no set standard interval for post-TARE imaging, however, common practice is imaging at 1 month after the procedure, 3 months after the first follow-up scan, and then every 3–6 months afterwards [8].

2.6 Current applications of TARE

TARE has shown a multitude of benefits in various clinical scenarios which include curative therapy, bridging to transplant, and palliation. As mentioned earlier, the two main primary methods of TARE treatment include segmental or lobar treatments. Segmental infusions, described in literature as “radiation segmentectomy” is a growing area of clinical interest due to the ability for radiation dose escalation as Y90 microparticles are delivered to $\le$2 segments of liver while sparing the rest of the liver. Lewandowski et al. first compared the outcomes of glass Y90 radiation segmentectomy, in patients with solitary tumors $\le$5 cm, to other curative therapies such as surgical resection and radiofrequency ablation and found similar response rates, tumor control, and survival outcomes in tumors $\le$3 cm, highlighting the curative potential of this approach [28]. Salem et al. investigated the utility of glass Y90 TARE for solitary tumors $\le$ 8 cm in the multicenter LEGACY study [29]. The majority of patients received selective infusion and an 86.6% three-year overall survival rate was observed. Additionally, explanted livers showed complete pathological necrosis at doses greater than 400 Gy, establishing an ablative dose [29]. While TARE was initially thought to be an alternative to TACE in intermediate stage disease (BCLC stage B), the above-mentioned studies established the role of TARE in early-stage disease (BCLC stage 0 & A).

Lobar Y90 refers to treatment of an entire hepatic lobe via either the right or left hepatic artery. The dosage for lobectomy is usually lower than that of segmentectomy (120–150 Gy for glass microspheres), as there is a larger amount of radioactive microparticles delivered to the normal/non-tumorous surrounding liver parenchyma. A well-established phenomenon with high dose lobar Y90 is compensatory, contralateral hypertrophy of the remnant or untreated lobe [30]. This approach offers the advantage of local tumor control combined with a larger future liver remnant for those eligible for surgical resection. With the help of radiation lobectomy, the FLR (i.e. the untreated lobe) can increase by up to 45% of its original volume allowing for sufficient hepatic reserve to facilitate resection [31]. Lobar treatment has also shown benefit in advanced disease with portal vein tumor thrombus (PVTT) which is usually managed with systemic chemotherapy. Other LRTs including percutaneous ablation and TACE are not recommended for HCC with PVTT [7]. TACE is relatively contraindicated due to its embolic nature which could lead to hepatic infarction due to its embolic effect.

3. Conclusion

TARE is a safe and efficacious LRT that plays a large role in the treatment paradigm for HCC with applications in curative therapy, bridging to transplant, as well as palliation. TARE is a two-stage procedure that includes both a mapping procedure as well as the Y90 microparticle infusion. The mapping procedure informs operators on the vascular anatomy of the tumor as well as variant hepatic arterial anatomy, arteriovenous shunting, and hepatoenteric/hepatogastric collaterals. There are two general methods of TARE which include radiation segmentectomy and radiation lobectomy. Segmentectomy has outcomes similar to other curative procedures while lobectomy has shown utility in patients undergoing resection as well as in portal vein invasion.

Acknowledgments

We thank Omar Ahmad, Medical Student.

Conflict of interest

O.A. has received speaking fees from Argon Medical, Cook Medical, Cardiva Medical, and Canon Medical and has served on an advisory board for Boston Scientific and Genentech.