INTRODUCTION
Hepatocellular carcinoma (HCC) claims half a million lives across the globe each year[1]. It is the sixth most common cancer in the world and is the third most common cause of cancer-related mortality[2]. Various etiologic factors have been implicated in the transformation of benign hepatic parenchyma to malignancy, however, no one factor has been shown to cause cancer in all cases. Although several postulates for tumorigenesis have been proposed, the exact underlying mechanism for neoplastic change remains unknown[13].
The incidence of HCC varies considerably across geographic regions with some areas reporting cases as high as 20/100 000 per annum[3]. Various studies have shown that advanced age and male sex portends a higher likelihood of developing HCC. Several important risk factors have been identified which substantially increase the possibility of developing disease. Among these, the most common risk factor recognized worldwide is the hepatitis B carrier state. Others inciting factors include chronic hepatitis C infection, cirrhosis, environmental toxins such as aflatoxin and contaminated drinking water, alcohol abuse, diabetes mellitus, and hereditary hemochromatosis.
In patients diagnosed with this lethal malignancy, less than 15% are candidates for surgical procedures. A survival benefit has been observed in patients that meet the rigorous criteria for curative resection or transplantation[4]. For the remaining majority, various treatments options have become available without universal agreement on which treatment option offers the greatest survival benefit with the least toxicity.
The use of external beam irradiation has historically played a limited role in the treatment of HCC due to the radiosensitive nature of normal hepatic tissue[5]. Investigators have shown that liver exposure to radiation doses greater than 40 Gy may result in a clinical syndrome characterized by ascites, anicteric hepatomegaly, and elevated liver enzymes weeks to months following therapy[56]. Additionally, a condition recognized as veno-occlusive disease, marked by central venous congestion and atrophy of adjacent hepatocytes, may develop. Together, the clinical and pathologic spectrum described above has been referred to as radiation induced liver disease (RILD) or radiation hepatitis. This is the most prominent treatment related complication in patients undergoing hepatic irradiation from external sources[7].
Given this limitation and the need for higher doses to inflict lethal injury to malignant tissue[8–10], minimally invasive intra-arterial devices have emerged. These devices, loaded with radioactive Yttrium-90 microspheres or Rhenium-188, can deliver very high tumoricidal doses without the development of RILD[11]. Using segmental infusion techniques, doses as high as 4993 Gy to liver tissue have been reported[12]. Although the use of these devices dates back to the early 1960’s, only recently has the therapeutic safety and efficacy associated with its use been realized[1314]. For the purpose of this review, we aim to highlight the use of intra-arterial radiotherapy for the treatment of inoperable HCC and update the reader on recent clinical and research advancements.
DEVICE AND DOSIMETRY CONSIDERATIONS
Yttrium-90 intra-arterial radiotherapy, also known as radioembolization, is a minimally invasive catheter-based therapy that delivers internal radiation via the arterial vessels that feed tumors. “Radio” refers to the radiation that is imparted to tissue; “embolization” refers to the microembolic effect[15]. This technology takes advantage of the dual blood supply to the liver. Normal hepatic tissue derives greater than 70% of its blood supply by way of the portal system whereas malignant tissue is preferentially supplied by the arterial system. There are currently two commercially available Yttrium-90 microsphere devices. TheraSphere® (MDS Nordion, Ottawa, Ontario, Canada) is made of glass and SIR-Spheres® (Sirtex Medical, Sydney, Australia) is made of resin. These two devices are different in a number of important respects[16]. TheraSphere® is a minimally embolic device consisting of 20-30 micron particles with higher specific activity (2500 Bq) and lower number of spheres (1.2 million microspheres/3 GBq). Conversely, SIR-Spheres® are moderately embolic, consisting of 20-60 micron particles, with lower specific activity (50 Bq), and greater number of spheres (approximately 40-80 million spheres/3 GBq). A third agent, available in certain countries, uses a Rhenium-188 based radioconjugate delivered in a trans-arterial manner analogous to the Yttrium-90 based devices[14].
90Y Glass microspheres
90Y microspheres (TheraSphere®, MDS Nordion, Ottawa, Canada) are composed of nonbiodegradable glass microspheres ranging from 20 to 30 &mgr;m in diameter, in which 90Y is an integral constituent of the glass. 90Y is a pure β-emitter with a physical half-life of 64.2 h, after which 90Y decays into stable zirconium. The average energy of β-emission is 0.9367 MeV, mean tissue penetration of 2.5 mm and a maximum penetration of 10 mm. One gigabecquerel (27 mCi) of 90Y per kilogram of tissue provides a dose of 50 Gy. The microspheres are supplied in 0.5 mL of sterile, pyrogen-free water contained in a 0.3-mL V-bottom vial secured within a 12-mm clear acrylic shield. The specific activity is 2500 Bq at the time of calibration.
The typical method of calculating the required activity level (in GBq) to be injected and the actual dose delivered to the liver and lung has been previously published. CT or MR imaging is used to determine the targeted liver volume to be treated with 90Y microspheres[17–19]. The targeted liver volume is that portion of liver tissue that will be perfused once the catheter is in the desired location. A conversion factor of 1.03 g/cm3 is used to calculate the corresponding targeted liver mass from the targeted liver volume. The required activity is calculated from the following formula:
Activity (GBq) = [target dose (Gy) × target liver mass (kg)]/50
When lung shunt fraction (LSF) and percentage of residual activity (R) in the vial after treatment are taken into account, the actual dose delivered to the target mass is calculated by rearranging the previous equation as follows:
Dose (Gy) = [Infused activity (GBq) × 50 × (1- LSF) × (1-R)]/liver mass (kg)
Cumulative liver dose is defined as the accumulated dose to that specific volume that was treated multiple times. By targeting delivery to a hepatic segment or lobe, 90Y therapy results in high radiation doses to the tumor while sparing liver parenchyma. These tumoricidal doses have proven effective in the ability of 90Y microspheres to reduce tumor viability, demonstrating an increasing therapeutic effect with radiation dose[11].
90Y Resin microspheres
SIR-Spheres® consist of biodegradable resin-based microspheres containing 90Y. The average size of a sphere is 35 microns in diameter. Upon in vivo administration, the spheres are permanently implanted. Each vial contains 3 GBq of 90Y in a 5 mL vial. Each vial contains 40-80 million spheres. The activity per microsphere is 50 Bq at the time of calibration.
The radioactivity to the liver can be calculated by one of two methods:
(1) The first method allows the calculation be based on body surface area to determine an approximate tumor burden:
Activity (GBq) = body surface area (m2) - 0.2 + (% tumor burden/100)
(2) Based on a broad estimate of tumor burden which then requires the user to increase the recommended activity by 0.5 GBq per 25% increase in tumor burden.
Activity for tumor involvement < 25%, 25%-50% and > 50% are 2.0 GBq, 2.5 GBq and 3.0 GBq, respectively.
Using either dosimetry model, activity administered is decreased depending on the extent of identified lung shunt. Also, recent clinical practices have shown that an additional 25%-30% activity reduction is usually necessary for SIR-Spheres®[20]. The dosimetry model for SIR-Spheres® is based on whole liver infusion. If a lobar administration is intended, the activity to be administered should be calculated using whole liver volume and then corrected for the target volume anticipated for treatment. As an example, if a right lobe infusion is anticipated, the calculated GBq should be multiplied by the percentage of right lobe as a proportion to the entire liver. Dosimetric issues and technical considerations have been described in detail previously[162021].
Rhenium-188 radioconjugate
This is available through the use of a Rhenium-188 generator. The half-life of Rhenium-188 is 16.9 h. The isotope delivers high-energy beta (2.1 MeV max) and a low energy gamma (155 keV) emissions, permitting imaging. Usually, this radioconjugate is in the form of Rhenium-188 4-hexadecyl 1, 2, 9, 9-tetramethyl-4, 7diaza-1, 10-decaethaniol labeled with iodized oil. Dosimetry is based on the safe and tolerable dose to organs at risk including the liver, lungs and bone. A small scout dose of the radioconjugate (3.7 MBq) is administered on the day prior to treatment. Subsequently, transmission scans with a Rhenium-188 flood source are performed to determine the attenuation correction factors for lung and liver to be used in the dosimetric calculations the following day. Anterior and posterior images are obtained to calculate geometric mean counts. After correcting for scatter, regions of interest are placed around the whole liver, tumor and lungs. Using medical internal radiation dosimetry and by adjusting for the difference in total body and organ masses between the patient and the anthropomorphic model, the proper activity required is calculated using the following dose limitations: 12 Gy to the lungs, 30 Gy to the normal liver, 1.5 Gy to the bone marrow.
Absolute and relative contraindications
Two absolute contraindications exist for the use of 90Y microsphere treatment in any patient. The first includes a pretreatment 99mTc macro-aggregated albumin (MAA) scan demonstrating significant hepatopulmonary shunting that would result in > 30 Gy being delivered to the lungs with a single infusion or as much as 50 Gy for multiple infusions. The second includes the inability to prevent deposition of microspheres to the gastrointestinal tract with modern catheter techniques. A number of relative contraindications exist including non-compromised pulmonary function, adequate liver reserve, serum creatinine < 2.0 mg/dL, and a platelet count > 75 × 109/L. For relative contraindications, clinical judgment should be exercised when determining whether a patient is appropriate to undergo this procedure.
Observed toxicities
The most common clinical toxicity observed with the use of 90Y is a mild post-embolic syndrome. This syndrome, unlike that observed with other embolic treatments such as transarterial chemoembolization (TACE), includes fatigue, vague abdominal discomfort, pain, and fever[112223]. Other avoidable toxicities that occur as a result of non-target radiation include: cholecystitis, gastric ulceration, gastroduodenitis, pancreatitis, radiation pneumonitis, and RILD[1624–27]. With meticulous planning, careful selection, and proper technique, the majority of these toxicities can be mitigated. Finally, hematologic toxicities seen in the immediate post-procedural period include lymphopenia. This is not an unexpected finding given the sensitivity of lymphocytes to radiation. Despite this, no infectious complications have been documented[162128].
LITERATURE REVIEW
A comprehensive literature review was completed in 2006 describing the entire clinical and scientific evidence for 90Y in detail[13]. Since then, additional evidence has been generated[16]. A consensus panel report from the radioembolization brachytherapy oncology consortium concluded that there is sufficient evidence to support the safe and effective use of this loco-regional therapy in HCC patients[20]. The authors further suggested the need to investigate the benefits of 90Y in combination with other traditional therapies. The results from phaseIand II studies in combining 90Y with targeted therapies (Raf-kinase, EGFR) for HCC are underway and should provide valuable insight into the toxicity and efficacy of such regimens.
Sangro et al reported on 24 HCC patients with Child-Pugh A disease who underwent radioembolization with resin microspheres[29]. The median activity delivered was 2.2 GBq. The investigators reported a reduction in size of target lesions in 19 patients. Using RECIST criteria, 88% of the cohort had either partial response or stable disease. The authors did not observe any post-embolization syndrome and all patients were discharged within 24 h of treatment. Two patients became jaundiced at 1 mo and 3 mo after the procedure from uncertain causes. Two treatment-related deaths were recorded. At median follow-up of 12.5 mo none of the treated patients progressed. Given the tumor response and minimal toxicity profile, the investigators concluded that radioembolization is a viable therapy for patients with portal vein thrombosis and preserved liver function and that this therapy needs to be considered in patients who are awaiting transplant in order to prevent extension of disease beyond the Milan criteria.
Rivera et al presented a case report of a 42 years old hepatitis C cirrhotic male with tumor recurrence 22 mo post-transplantation[30]. The investigators in the study then treated the patient with 90Y resin microspheres. The author noted no change in liver function post-procedurally and follow-up MRI demonstrated the absence of arterial enhancement and tumor necrosis. The authors concluded the use of 90Y for post-transplant recurrence may help prolong patient and graft survival in patients that develop recurrence.
Gulec et al retrospectively analyzed the data from a heterogeneous cohort of 40 patients with primary and metastatic liver malignancies who underwent single whole liver treatments using 90Y resin microspheres[31]. The average administered activity was 1.2 GBq and tumor absorbed doses ranged from 40.1 to 494.8 Gy. Sixty-seven percent of the treated cohort responded to therapy with favorable responses reported in those with higher tumor flow ratios. The authors concluded that doses up to 100 Gy to the uninvolved liver were tolerated by this procedure without the development of veno-occlusive disease or liver failure. The authors further noted that lowest tumor dose necessary to generate a detectable response was 40 Gy.
Kamel et al reported on 13 patients prospectively treated with 90Y glass microspheres. MR imaging was compared 24 h pre-treatment to an average follow-up of 55 d post-therapy[32]. Targeted tumors demonstrated a mean decrease in arterial enhancement of 22%, a mean decrease in venous enhancement of 25% and unchanged tumor size in both targeted and non-targeted tumors. The authors reported a median survival of 12 mo from time of diagnosis and emphasized the need for surrogate imaging measures such as diffusion-weighted MR in order to assess response.
Keppke et al reported on the imaging findings and median survival of 42 patients using 90Y glass microspheres[33]. The response rates according to WHO, RECIST, necrosis and combined criteria (RECIST and necrosis) were 26%, 23%, 57% and 59%, respectively. The median survival for Okuda I patients was 660 d. The authors concluded that the imaging findings, using a combined criteria (size and necrosis), resulted in a more accurate assessment of tumor response after 90Y radiotherapy when compared to size criteria alone.
In an attempt to address the question of retreatment using this therapy, Young et al recently reported on the relationship between cumulative radiation dose and the development of liver toxicities in 41 patients stratified to OkudaIand II[34]. The authors observed a statistically significant mean cumulative radiation dose of 390 Gy and 196 Gy tolerated by OkudaIand Okuda II patients, respectively, before the occurrences of toxicity. This suggests that some patients can tolerate multiple treatments prior to the development of liver toxicities. Median survival from date of first treatment for OkudaIand Okuda II were 660 d and 431 d, respectively (P = 0.44).
More recently, Kulik et al reported on 21 patients from a large database of 251 patients who had undergone 90Y glass microsphere therapy and subsequently bridged to transplantation[35]. Target tumor dose administered was 120 Gy with toxicities including fatigue in the majority of patients (42%). The authors reported a mean reduction in alpha fetoprotein (AFP) of 33% from pre-treatment levels. The investigators noted complete necrosis by pathologic exam in 14 of 21 patients (66%). Four of 21 patients had disease recurrence with a mean time to recurrence of 250 d, a finding not uncommon following transplantation[36–38]. The authors concluded that treatment with 90Y achieves complete necrosis in the majority of targeted lesions in patients bridged to transplantation, but that recurrence was a possibility despite the radiographic findings of complete necrosis.
Additionally, Kulik et al reported on the safety of 90Y in a 108 patient cohort treated with glass microspheres, with subset analysis comparing patients with and without portal vein thrombosis[39]. Thirty-seven of 108 patients presented with imaging proven portal vein thrombosis (PVT). Patients were stratified by Okuda, Child Pugh, baseline bilirubin, ECOG, presence of cirrhosis and location of PVT (none, branch, and main). The cumulative dose administered to those with and without PVT were 139.7 Gy and 131.9 Gy, respectively. Liver related adverse events reported included elevation of bilirubin in 40%, ascites in 18%, and hepatic encephalopathy in 4% of the patients with cirrhosis and main PVT. In the patients without cirrhosis, elevated bilirubin occurred in 4%, ascites in 4% and no cases of encephalopathy. Tumor response using WHO criteria and EASL recommendations were 42.2 and 70%, respectively[40]. Median survival from the date of first treatment for patients without PVT and cirrhosis was 813 d. In patients with branch PVT, survival was 304 d from time of treatment (cirrhotics: 261 d, non-cirrhotics: 427 d). The authors concluded that the microembolic effect of 90Y microspheres did not increase the risk of liver adverse events in patients with proven PVT. Glass microspheres did not result in a microembolic effect that is seen with other loco-regional therapies using larger diameter particles.
Investigators also studied the use of Rhenium-188 for patients with inoperable HCC[14]. A multicenter clinical trial was completed looking at Rhenium-188 (Rh-188) lipiodol delivered in a transarterial manner. After complete clinical evaluation (including serum alpha-fetoprotein (AFP), tumor burden, patency of portal vein, Child-Pugh and Okuda classification), radiation absorbed dose (rad) to various organs, including tumor, was calculated after injecting 185 MBq of Rh-188 iodized oil via the hepatic artery. From this value, the maximum tolerable activity, defined as the amount of radioactivity delivering no more than 12 Gy of rad to lungs, 30 Gy to normal liver, or 1.5 Gy to bone marrow, was calculated and injected. Ninety-three patients were successfully treated with a mean age of 53 years (80 men and 13 women). Mean cumulative dose was 7.8 GBq. Sixty-eight percent of patients had serologic evidence of hepatitis B and/or C; 40% had clinical/radiologic evidence of cirrhosis. Mean tumor diameter was 10.3 ± 4.4 cm, with 40% of patients having more than three lesions; in 50% of patients, tumor was either unilateral, occupying 50% or more of the liver, or bilateral. AFP was elevated in 68% of patients and serum levels exceeding 300 ng/mL was observed in 44% of these patients. There was portal vein thrombosis in 38% of patients, Child-Pugh B disease in 37% of patients, and Okuda stage II or III disease in 50% of patients. Mean first administered activity was 5.3 GBq ± 1.6, which delivered 88 Gy to the tumor. Treatment was tolerated well. Five patients had complete tumor response, while 17 had a partial response (> 50% tumor reduction) for an overall objective response rate of 33%. Thirty-five percent of patients had stable disease. Only dose to the tumor was found to be significantly (P = 0.001) associated with tumor and/or AFP response. Median survival for the entire cohort was 356 d and varied accordingly with baseline characteristics. Responders by imaging survived longer than those that did not exhibit a response and interestingly was correlated with dose administered to tumor. The authors concluded that this was a safe, effective, and promising therapy in patients with HCC with favorable cost-benefit profile.
