Complications Following Radioembolization with Yttrium-90 Microspheres: A Comprehensive Literature Review

Primary Tumors 

THE most common primary malignancy of the liver is hepatocellular carcinoma (HCC); its incidence is increasing worldwide. It ranks as the sixth most common malignancy and third most common cause of cancer-related mortality (1, 2). The management of liver tumors has seen significant advancement in the past decade with the development of new screening, diagnostic, and therapeutic modalities. Primary liver tumors include HCC and intrahepatic cholangiocarcinoma. Surgical resection is considered curative in patients with resectable HCC and normal liver function (3). Transplantation is considered the gold standard for patients with HCC whose HCC meets Milan criteria (4). The role of transplantation is nevertheless limited given few available donor organs. Tumor progression may also render patients ineligible for transplantation given increasing wait times (5, 6, 7). Transarterial chemoembolization and radiofrequency ablation represent standard therapies in treating selected patients and serving as a “bridge” to transplantation (8, 9). Radioembolization to delay disease progression has an emerging role as a bridging therapy in cases of HCC within Milan criteria. It has also been shown to downstage HCC tumors that do not meet Milan criteria to meet the criteria for transplantation (10, 11). Patients with unresectable disease as a result of tumor multifocality or macrovascular tumor involvement have also exhibited evidence of clinical benefit after radioembolization (12).

Secondary Tumors 

The liver is a common site of malignancies, the most common being metastatic from other primary tumors (13). Numerous primary malignancies such as colorectal carcinoma metastasize to the liver, as it serves as a filter between the portal venous flow from the gastrointestinal (GI) tract and the hepatic venous system. Secondary liver tumors are managed by surgical and medical treatments. The role of radioembolization for secondary liver tumors is promising, and it has been shown to be safe and efficacious in patients with colorectal carcinoma, neuroendocrine tumors, and other polychemotherapy-refractive conditions (14, 15, 16).

Pretreatment Evaluation 

HCC is diagnosed on imaging if there is a lesion greater than 2 cm with arterial-phase enhancement and venous washout (17). Biopsy is performed in lesions that do not have the typical radiologic findings as defined by the European Association for the Study of the Liver and American Association for the Study of Liver Diseases guidelines (18). The role of α-fetoprotein in diagnosis of HCC is not yet established. Secondary liver tumors are diagnosed by imaging with fluorodeoxyglucose positron emission tomography and conventional anatomic imaging modalities such as computed tomography (CT). Appropriate laboratory tests including a hepatic panel and corresponding tumor marker measurements are performed to ascertain baseline values. A detailed chemotherapeutic drug history is also important as chemotherapy may predispose a patient to complications that may be seen after yttrium-90 (90Y) therapy. Ideally, patients are selected for radioembolization by a multidisciplinary team that should consist of hepatologists, medical and surgical oncologists, transplant surgeons, and interventional radiologists.

Pretreatment Angiography 

It is necessary to perform mapping angiography before radioembolization as this provides the interventional radiologist with knowledge of the hepatic arterial anatomy. This is a crucial step in minimizing adverse events. The presence of atherosclerosis in the aorta and its tortuosity is assessed based on the aortogram. The superior mesenteric angiogram is used to look for variant vessels to the liver (eg, accessory or replaced right hepatic artery) and is also used to assess the patency of the portal vein. The celiac artery angiogram assesses the hepatic vasculature and variants. Selective left and right hepatic artery angiography are performed to assess segmental flow. The gastroduodenal arteriogram is obtained to analyze the flow to the stomach, duodenum, and pancreas. The caudate lobe (segment 1) requires special attention. One should determine the blood flow to segment 1 tumors by identifying all contributing feeder vessels (eg, left hepatic artery, right hepatic artery, right inferior phrenic artery, right renal artery, right adrenal artery, aorta), given the caudate lobe as an anatomic watershed area (19).

Coil Embolization 

Coil embolization of the communications between the liver and GI vessels is an important prerequisite of treatment. The aberrant deposition of microspheres in the GI tract or pancreas can have grave consequences. Coil embolization is a safe and efficacious mode of preventing hepaticoenteric flow.

Prophylactic embolization of the gastroduodenal artery (GDA) and right gastric artery is therefore recommended, particularly if resin microspheres are used (20, 21, 22). In cases of large peripherally located tumors, omental collateral vessels arising from the GDA may supply the tumor. Permanent occlusion of the GDA may prevent further embolic therapy to the tumor via the omental vessels. The degree of pretreatment prophylactic embolization before radioembolization should be based on the treating physicians' experience, vessel size, planned treatment location, and radioembolic device being considered (ie, glass vs resin) (23, 24).

The following are important variant vessels that may require prophylactic coil embolization: falciform artery, right or left inferior phrenic artery, inferior esophageal artery, right or accessory gastric artery, supraduodenal artery, retroduodenal artery, and accessory right hepatic artery (from GDA) (20, 23, 24).

Technetium-99m Macroaggregated Albumin (99mTc-MAA) Scan 

A 99mTc-MAA scan is performed to assess the lung shunt fraction (LSF) and splanchnic shunting.

Yttrium-90 

90Y is a pure β-emitter. It has an average β-energy emission of 0.9367 MeV and a maximum of 2.1 MeV. The mean tissue penetration is 2.5 mm with a maximum of 10 mm.

TheraSphere 

TheraSphere particles (MDS Nordion, Ottawa, ON, Canada) are glass microspheres 20–30 μm in size with 90Y as anintegral constituent (26). One 3-GBq vial contains 1.2 million particles. Given the low particle number, the particles impart a minimal embolic effect (27). This device was approved by the Food and Drug Administration for use in treatment of unresectable HCC in 1999. The activity per microsphere is 2,500 Bq at calibration. There are six vials of different activities available calibrated for Sunday at 12:00 p.m. Eastern Standard Time (28). Dosimetric considerations have been described in detail elsewhere (23, 29).

SIR-Spheres 

SIR-Spheres (Sirtex Medical, Lane Cove, Australia) are composed of resin microspheres with sizes ranging from 20 to 60 μm. One 3-GBq vial contains 40–80 million particles. Given the particle number, the particles impart a moderate embolic effect. This device was approved by the Food and Drug Administration for use in colorectal carcinoma in 2002 (30). The activity per microsphere is 50 Bq at calibration. One 3-GBq vial is provided by the manufacturer calibrated for 6 p.m. on the day of treatment. Dosimetry for resin microspheres has also been described previously (29, 31).

Summary of PRS 

The incidence of PRS in the literature ranges from 20% to 55% (11, 21, 32, 33). The duration of PRS has not been studied. Patients may be given steroids and antiemetic agents to minimize the incidence of PRS. A follow-up with the patient 2 weeks after radioembolization is recommended to inquire for clinical evidence of PRS. Symptomatic management may be required.

Hepatic Dysfunction 

Young et al (34) studied 41 patients who underwent 90Y radioembolization, all of whom had the same liver tissue irradiated multiple times. This study was performed to determine the safe radiation dose limits to the liver for treatment of HCC. A conservative approach was used to determine the toxicity to the liver. An increase in bilirubin, alkaline phosphatase, alanine aminotransferase, or aspartate aminotransferase, or a decrease in albumin, was considered to indicate a derangement of liver function. A total of 13 toxicities were observed in seven patients. It was noted that patients with Okuda stage I disease could tolerate a dose as high as 390 Gy whereas those with Okuda stage II disease could tolerate a dose as high as 196 Gy of radiation. There was a selection bias in this study, as only patients well enough to undergo repeat treatment were analyzed. The presence of cirrhosis was a significant confounding variable when assessing liver toxicities in patients with HCC (34).

Sangro et al (35) also studied the liver disease induced by radioembolization of liver tumors in 45 patients who underwent the treatment for primary or secondary liver tumors. Nine patients (20%) developed clinical complications such as jaundice and ascites 4–8 weeks after therapy (35). The histologic hallmark of venoocclusive disease was seen in severe cases, but not in patients who were chemotherapy-naive. The incidence of radiation-induced liver disease (RILD) was associated with increasing age, whole-liver treatment, and increased baseline bilirubin levels. RILD can produce significant morbidity and may be potentially life-threatening.

Kennedy et al (36) recently studied the incidence of RILD after 680 90Y treatments with resin microspheres. RILD was observed after 28 treatments (4%). Their data suggest an association between the activity delivered to the patient and RILD (36). There may also be an association between the use of the empiric method for the calculation of the dose (for resin spheres) and toxicity. Twenty-one of the 28 cases of RILD observed were from one center that used the empiric method for dosimetry.

Summary of Hepatic Dysfunction 

The incidence of RILD after 90Y administration ranges from 0%–4% (34, 35, 36). RILD results from the exposure of normal liver parenchyma to high doses of radiation. This may lead to biochemical aberrations with minimal clinical manifestations; clinical correlation is essential. Follow-up laboratory evaluation is routinely recommended at 1 month after treatment. In cases of clinically manifest RILD, supportive management is recommended. A biopsy of the normal parenchyma may help confirm the diagnosis of RILD.

RILD is seen most often in patients with preexisting liver function abnormalities. Patients with baseline bilirubin levels more than 2 mg/dL are generally not considered ideal candidates. Whole-liver radioembolization in a single session is not recommended. The dosimetry for radioembolization is complex and the authors do not recommend the use of the empiric method for dose calculation when using SIR-Spheres.

Biliary Sequelae 

Atassi et al (37) studied the biliary sequelae after radioembolization with 90Y in 327 patients with primary or secondary liver tumors. Biliary sequelae were evaluated at 3 months with conventional anatomic imaging techniques such as CT or magnetic resonance imaging. Thirty-three patients were found to have 40 imaging findings related to the biliary tree. These findings included biliary necrosis (n = 17), bilomas (n = 3), cholecystitis (n = 2), gallbladder wall enhancement (n = 6), gallbladder wall rent (n = 3), abscess (n = 1), and biliary strictures (n = 8). Figure 1 illustrates biloma formation after radioembolization. Figure 2 illustrates an abscess formed after radioembolization. Six of these patients (1.8%) needed unplanned surgical or interventional procedures to treat the complication. Radiation-induced cholecystitis in certain settings may require cholecystectomy. Histologic analysis of the gallbladder after cholecystectomy has shown microspheres in the gallbladder wall. Figure 3 depicts radiation cholecystitis. Because many of the imaging findings suggestive of biliary sequelae do not have clinical consequences, clinical correlation with the imaging findings are necessary (38, 39).

View Large Image Download to PowerPoint

Figure 1. (a) Posttreatment contrast-enhanced CT in the arterial phase demonstrates a fluid collection representing a small biloma (arrow). (b) Coronal view in the same patient confirms the presence of the biloma (arrow). With kind permission from Springer Science & Business Media: Abdominal Imaging, Radiologic findings following y90 radioembolization for primary liver malignancies, in press, Ibrahim SM, Nikolaidis P, Miller FH, et al.

View Large Image Download to PowerPoint

Figure 2. Contrast-enhanced posttreatment CT of metastatic pancreatic cancer to the right hepatic lobe in a patient with a history of Whipple surgery. Focal, peripheral enhancing lesion (white arrow) containing fluid and air, suspicious for hepatic abscess, is seen. The patient was successfully treated with percutaneous drainage and antibiotics. Reprinted with permission from Atassi B, Bangash AK, Bahrani A, et al, RadioGraphics, 2008, 28:81–99, by Radiological Society of North America.

View Large Image Download to PowerPoint

Figure 3. Contrast-enhanced CT 4 weeks after 90Y microsphere radioembolization in a 45-year-old woman with breast cancer metastatic to the liver (a) shows radiation-induced disruption and discontinuity in the enhancement of the gallbladder wall. This patient required cholecystectomy. (b) Hemorrhage within the gallbladder wall. A microsphere is seen (black arrow). (c) CT in a patient with pancreatic cancer metastatic to the liver shows gallbladder wall edema and disruption (white arrow) 1 month after 90Y treatment. Despite this concerning imaging finding, this patient had no symptoms of cholecystitis and did not require intervention. Reprinted with permission from Atassi B, Bangash AK, Bahrani A, et al, RadioGraphics, 2008, 28:81–99, by Radiological Society of North America.

Ng et al (40) found two unique biliary complications after radioembolization. One patient who was treated with 90Y for HCC presented with obstructive jaundice and was found to have biliary strictures. The other patient, who received radioembolization to treat a secondary liver tumor, presented with fever, jaundice, and right upper quadrant pain and was found to have radiation-induced cholangitis. The authors suggested the use of liver biopsy to confirm the diagnosis of radiation-induced biliary complications in select settings.

90Y glass microspheres in the setting of tumor-related biliary obstruction (ie, without violated ampulla or stent placement) have been demonstrated to have a good safety profile without evidence of progressive leukocytosis, bilirubin increase, or biliary complications after radioembolization (41).

There is a protective effect against biliary complications that exists in patients with cirrhosis with HCC (42). Cirrhosis leads to hypertrophy of the peribiliary capillary plexus, and as a result, patients with cirrhosis have a lower incidence of biliary sequelae versus those without cirrhosis.

Summary of Biliary Sequelae 

The incidence of biliary sequelae after radioembolization is less than 10% (38, 39, 40). These complications may result from the microembolic effect of the therapy or radiation-induced injury to the biliary structures. Most biliary complications are not manifest clinically; clinical correlation with imaging findings is recommended. Abscesses may require drainage and antibiotics. Radiation cholecystitis requiring surgical intervention occurs in less than 1% of cases (38).

Radiation cholecystitis may be prevented by identifying the cystic artery and injecting microspheres distal to its origin. If blood flow into the cystic artery is significant and radioembolization distal to its origin is not possible, embolization may be considered (43). Patients with metastatic disease with a history of systemic polychemotherapy may also be at high risk of developing biliary complications.

Portal Hypertension 

Ayav et al (44) presented a case report in 2005 of a patient with colorectal carcinoma who had undergone chemotherapy for primary colorectal cancer and underwent a left lateral hepatectomy for liver metastases. The patient later underwent radioembolization of the right lobe. Given the dramatic imaging response, a curative liver resection of the liver was planned. The resection was not possible as a result of mesenteric portal hypertension and bleeding during surgery (44). Liver biopsy revealed fibrosis, which was not present in the excised left lobe, establishing 90Y as the culprit.

Jakobs et al (45) presented their analysis of fibrosis, portal hypertension, and hepatic volume changes induced by radioembolization. A total of 32 patients with metastatic disease to the liver were identified. Patients with secondary liver metastases (as opposed to HCC) were selected as the population to study so the confounding variable of cirrhosis could be excluded when attributing the adverse event of portal hypertension to radioembolization (45). The volumes of the spleen and whole liver and its lobes were measured before and after treatment. The diameters of the superior mesenteric vein, splenic vein, and main, right, and left portal vein were also measured before and after treatment. Patients who were receiving chemotherapy before treatment had preexisting portal hypertension based on the presence of chemotherapy-associated steatohepatitis. The mean decrease in liver volume was 11.8% and the mean increase in splenic volume was 27.9% in patients who had undergone bilobar treatment. Unilobar treatment did not lead to portal hypertension as the nontreated lobe was able to compensate for the atrophy of the treated lobe with contralateral lobe hypertrophy. The authors did conclude that radioembolization may cause portal hypertension based on imaging (45). None of the patients exhibited any clinical sequelae of portal hypertension.

Summary of Portal Hypertension 

Despite the imaging findings indicative of portal hypertension, the clinically significant occurrence of portal hypertension is low (45). Radiation leads to fibrosis, which causes the hepatic parenchyma to contract. This can radiologically manifest itself as portal hypertension. However, clinically relevant manifestations such as reduced platelet counts (<100,000/dL) or variceal bleeding are rarely seen. It is recommended to observe for radiologic and clinical evidence of portal hypertension routinely, as this is not an acute process (46).

A majority of patients with HCC have portal hypertension on imaging (ie, splenomegaly) that results from cirrhosis. However, this finding in isolation is not a contraindication to radioembolization.

Radiation Pneumonitis 

Leung et al (47) presented their data on 80 patients treated with 90Y for hepatic tumors. Five of these patients (6.3%) developed a restrictive ventilatory dysfunction; this was labeled radiation pneumonitis. All patients who developed radiation pneumonitis had LSFs greater than 13%. Resin microspheres were used in this study; these microspheres are known to have a combined radiation and embolic effect. Radiation pneumonitis can be seen as a typical “bat-wing” appearance on chest CT (47).

Salem et al (48) presented data that showed minimal pulmonary complications after radioembolization. A total of 403 patients with liver tumors were treated with radioembolization with glass microspheres. A cumulative radiation dose of greater than 30 Gy was delivered to the lung in 58 patients. Fifty-three of these 58 patients had follow-up lung imaging. Only 10 of these patients had imaging findings related to pulmonary complications, ie, pleural effusion, atelectasis, or ground-glass attenuation. There was no clinical or radiologic evidence of radiation pneumonitis. The Radiation Therapy Oncology Group/European Organisation for the Research and Treatment of Cancer criteria for radiation pneumonitis (49) (Table 1) were used, and 10 patients (19%) had grade 1 toxicities. Toxicities of a higher grade were not seen. The authors concluded that limitations of lung dose from radioembolization with glass microspheres (ie, 30 Gy/treatment, 50 Gy cumulative) need to be redefined, as these doses were well tolerated. Figure 4 represents an example of low and high LSF as seen on a 99mTc-MAA scan.

Table 1. Pulmonary Toxicity Grading Scale: Radiation Therapy Oncology Group/European Organisation for the Research and Treatment of Cancer

	Grade	Clinical/Radiologic Findings
	0	None
	1	Asymptomatic/mild symptoms (cough), slight radiographic appearance
	2	Moderate symptomatic fibrosis/pneumonitis (severe cough/fever), patchy radiographic appearance
	3	Severe symptomatic fibrosis/pneumonitis, dense radiographic appearance
	4	Respiratory insufficiency requiring, continuous O2/assisted ventilation
	5	Death directly related to radiation effects

View Large Image Download to PowerPoint

Figure 4. (a) 99mTc-MAA scan on the left shows an LSF of 0% and 99mTc-MAA scan on the right shows a low LSF. (b) 99mTc-MAA scan shows high LSF (arrowheads indicate lung uptake).

Summary of Radiation Pneumonitis 

The incidence of radiation pneumonitis if standard dosimetry models are used is well below 1% (47, 48). Radiation pneumonitis manifests as a restrictive ventilatory dysfunction. It is radiologically seen as a bat-wing appearance on chest CT. Management is medical; steroids may play a role. Other complications such as atelectasis and pulmonary effusion may be rarely seen.

The 99mTc-MAA scan is essential to calculate the LSF. A high LSF translates into a high percentage of the activity (and hence dose) being delivered to the lungs. Lung doses less than 30 Gy per treatment and less than 50 Gy cumulatively are recommended.

GI Complications 

Carretero et al (50) presented their data on 78 patients who were treated with 90Y radioembolization. Gastric and duodenal injury was seen in three patients (4%). The clinical presentation of this injury was intense pain during or after the procedure. Upper endoscopy showed wide areas of ulceration. The authors concluded that radioembolization carried an inherent risk of inducing GI ulceration secondary to misdirected spheres.

Murthy et al (51) presented data on the GI complications associated with radioembolization and attributed these toxicities to unrecognized variants, collateral circulation, and changes in flow dynamics during infusion. The information from the pretreatment angiography assists in minimizing this complication. Two patients were seen to have GI ulceration after radioembolization. Both patients were found to have an aberrant right gastric artery arising from the left hepatic artery that was not identified before the procedure. Upper endoscopy identified GI ulceration. Microspheres were visualized on histologic examination of the biopsy specimen of the ulcer. The nuclear scans have limited utility in assessing splanchnic shunting and must be correlated with angiographic findings. The use of prophylactic gastric acid suppressive agents was advised after treatment.

Mallach et al (52) reported a case of gastroduodenal ulceration after radioembolization. The patient had sigmoid adenocarcinoma metastatic to the liver. He presented with epigastric pain, nausea, and anorexia after treatment, which was refractory to medical management. Upper endoscopy showed a wide area of ulceration and the biopsy showed microspheres within the gastric arterioles.

Szyszko et al (53) described one patient who developed an ulcer in their analysis of 21 patients. Neff et al (54) had a similar study that showed the toxicity of radioembolization and reported GI ulceration in 29% of their 21 patients. This unacceptably high rate of ulceration may be explained in part by the six patients who underwent radioembolization via the proper hepatic artery; this approach is not recommended by guidelines. A recent study by South et al (55) showed GI ulceration in three of 27 patients (11%) treated with radioembolization. They concluded that refractory ulcers should be considered for aggressive surgical management.

Summary of GI Complications 

The incidence of GI ulceration is less than 5% if proper percutaneous techniques are used (51, 52, 53). The pathophysiology behind this complication is the ectopic distribution of radioembolic microspheres into the lining of the GI tract. Figure 5 shows a GI ulcer that occurred after radioembolization. Severe epigastric pain after treatment should be aggressively managed as early management could prevent more serious complications from ensuing. Endoscopy may be required to confirm the diagnosis. Cases refractory to proton pump inhibitors may require surgical management. As opposed to a normal ulcer that develops at the mucosal surface, 90Y-induced ulcers originate from the serosal surface. This may theoretically decrease the ability of the ulcer to heal and complicate the surgical field from scar/adhesions should surgery be required.

View Large Image Download to PowerPoint

Figure 5. (a) Contrast-enhanced CT scan of a patient with multifocal cholangiocarcinoma treated with 90Y. The patient developed a gastroduodenal ulcer after therapy. Note air in the ulcer and a thickened pyloric wall (arrow). (b) Coronal contrast-enhanced CT scan in the same patient demonstrates air in the ulcer and a thickened pyloric wall (arrow). (c) This patient had confirmed ulceration at endoscopy (arrow). With kind permission from Springer Science & Business Media: Abdominal Imaging, Radiologic findings following y90 radioembolization for primary liver malignancies, in press, Ibrahim SM, Nikolaidis P, Miller FH, et al.

Pretreatment angiography is essential to identify vessels that may supply the GI tract (20, 21, 22). Prophylactic embolization of the GDA is recommended if a high number of microspheres are to be delivered. The right gastric artery may come off the proper hepatic artery and may require embolization. The left hepatic angiogram is obtained to identify left gastric, inferior esophageal, and right gastric arteries. Prolonged and delayed angiography of the left hepatic artery is recommended; opacification of the coronary vein confirms gastric or esophageal flow. The right hepatic angiogram is required to identify the supraduodenal and retroportal arteries. The 99mTc-MAA scan may show splanchnic flow but must be correlated to angiographic findings (25). Prophylactic gastric acid suppressive agents are recommended after therapy.

As stated earlier, the degree of pretreatment prophylactic embolization should be determined based on the treating physicians' experience, vessel size, planned treatment location, and radioembolic device being considered (23, 24). The need for detailed angiography and proper angiographic technique based on accepted and published standards cannot be under-emphasized. These authors believe that most, if not all, gastrointestinal toxicities can be avoided by using meticulous technique.

Vascular Injury 

Murthy et al (56) presented their data on radioembolization in 10 patients who were being treated with the chemotherapeutic agents cetuximab or bevacizumab in addition to resin microspheres. The data present a reasonable safety profile with the use of these drugs but advocates further investigation.

Chemotherapy has been shown to make the vasculature more friable and prone to injury. There have been cases of dissection and rupture of the vessels despite adherence to normal protocol in these patients. An analysis of 16 patients receiving chemotherapy (57) demonstrated abnormalities in the vasculature and hepatic arterial flow in 12 (75%). During angiography, a thorough search for stenoses, aneurysms, and flow abnormalities should be undertaken, particularly if the patient has been exposed to chemotherapy (57). Figure 6 illustrates vascular dissection (despite use of microcatheters) in a patient receiving previous systemic chemotherapy for colon cancer.

View Large Image Download to PowerPoint

Figure 6. (a) Pretreatment angiogram in a patient who had received previous systemic chemotherapy for colorectal carcinoma. (b) Posttreatment angiogram shows vascular dissection (arrow) and distal perfusion defect (arrowhead). This patient underwent successful management with antiplatelet agents.

Summary of Vascular Injury 

Although the incidence of vascular injury during radioembolization is low, it is seen most often in patients receiving chemotherapy. In the case of arterial dissection, angioplasty/stent placement or antiplatelet agents may be required. The use of microcatheters and careful wire/catheter manipulation is recommended in patients receiving (or previously exposed to) systemic chemotherapy.

Lymphopenia 

Lymphopenia is a possible clinical sequela of 90Y infusion. Lymphocytes are extremely radiosensitive and lymphocyte toxicity may be seen after glass microsphere radioembolization. Greater than a 25% decrease in lymphocyte count after treatment is seen in the majority of patients (10, 11). There have been no reports of opportunistic infections resulting from lymphopenia after radioembolization (10, 11).

Miscellaneous Adverse Events 

Periumbilical pain has been reported and may be a result of inadvertent spread of microspheres to the vessels supplying the anterior abdominal wall via the falciform artery (20, 23, 24). In fact, radiation dermatitis from nontarget microsphere flow into the falciform artery has recently been described (58). Pretreatment angiography should help identify the falciform artery. Prophylactic embolization of this vessel is recommended if possible to mitigate abdominal pain.

There are rare and unusual side effects that may be seen after treatment with 90Y. Patients have been known to experience acute chills lasting minutes during treatment (29). Such cases usually respond to diphenhydramine and meperidine. Gustatory abnormalities including a transient metallic taste have also been reported. The adverse events associated with all diagnostic and therapeutic angiographic procedures, such as hematoma formation at the puncture site, may also be seen in radioembolization (59).