Residency · Residency · Interventional Radiology

Transarterial Radioembolization (TARE) with Yttrium-90

Principles of Radioembolization

Selective internal radiation therapy (SIRT): delivery of radioactive microspheres into the hepatic arterial supply of liver tumors. Microspheres lodge in the tumor microvasculature and deliver high-dose beta radiation (Yttrium-90) locally. Y-90 is a pure beta emitter with a mean tissue penetration of 2.5 mm (max 11 mm) and a half-life of 64.1 hours. 94% of radiation is delivered within 11 days. Unlike TACE, radioembolization is NOT embolic at the doses used -- it is a brachytherapy, not an embolic procedure.

Available Products

Glass microspheres (TheraSphere): 20-30 microns; higher specific activity (2,500 Bq per sphere); fewer spheres per treatment (1-5 million); less embolic effect. Resin microspheres (SIR-Spheres): 20-60 microns; lower specific activity (50 Bq per sphere); more spheres per treatment (20-80 million); slightly more embolic effect. Both are FDA-approved; choice is often institutional.

Indications

Hepatocellular carcinoma: BCLC B (alternative to TACE), BCLC C with portal vein thrombosis (advantage over TACE which is contraindicated with main PVT), bridge/downstaging to transplant or resection. Colorectal liver metastases: salvage therapy after failed systemic chemotherapy; increasingly combined with chemotherapy. Neuroendocrine tumor liver metastases: symptomatic control and tumor debulking. Intrahepatic cholangiocarcinoma: emerging indication; limited but growing evidence. Other metastases: breast, melanoma, and other hypervascular liver-dominant metastatic disease.

Preprocedural Workup

Mapping Angiography (Planning Arteriogram)

Performed 1-2 weeks before Y-90 treatment. Selective hepatic arteriography to map the tumor-feeding arteries and identify: Variant hepatic artery anatomy. Extrahepatic arteries that could lead to non-target radiation (gastroduodenal, right gastric, cystic, falciform, phrenic arteries). Coil embolization of non-target vessels at the time of mapping (GDA, right gastric artery, etc.) to prevent non-target microsphere deposition. Prophylactic coiling is essential to prevent radiation-induced gastritis/ulceration.

MAA Scan (Technetium-99m Macroaggregated Albumin)

After mapping, inject Tc-99m MAA into the same arterial position planned for Y-90 delivery. MAA particles (10-90 microns) simulate the distribution of Y-90 microspheres. SPECT/CT imaging obtained to assess: Intrahepatic distribution (confirms tumor targeting). Extrahepatic deposition (stomach, duodenum -- indicates need for additional coil embolization). Lung shunt fraction (LSF): percentage of MAA that passes through arteriovenous shunts in the tumor into the pulmonary circulation. LSF <10%: no dose modification needed. LSF 10-15%: may require dose reduction. LSF >15-20%: relative contraindication (risk of radiation pneumonitis). Dose to the lungs should not exceed 30 Gy per treatment or 50 Gy cumulative.

Dosimetry

BSA Method (Body Surface Area)

Empiric dosing based on patient body surface area and percent tumor involvement. Simple but does not account for individual tumor biology or liver volume. Formula: Activity (GBq) = BSA - 0.2 + (% tumor involvement / 100). Standard for resin microspheres.

Partition Model (MIRD-Based)

Divides the liver into tumor, normal liver, and lung compartments. Calculates the dose (Gy) to each compartment based on MAA uptake ratios. Allows personalized dosing to maximize tumor dose while limiting normal liver dose. More complex but more accurate; preferred for TheraSphere. Target tumor dose: >120 Gy for HCC (higher doses correlate with better response).

DOSISPHERE Concept

Personalized dosimetry using partition model targeting >205 Gy to the tumor. DOSISPHERE-01 trial: personalized high-dose Y-90 showed significantly better objective response rate (71% vs. 36%) and overall survival compared with standard dosimetry in HCC. Paradigm shift toward personalized dosimetry in Y-90.

Treatment Procedure

Arterial access (femoral or radial). Catheter positioned in the same location as the MAA injection. Y-90 microspheres administered slowly under fluoroscopic guidance (check for reflux). Stasis is NOT the endpoint (unlike TACE); deliver the prescribed dose. Post-delivery angiography to confirm catheter position unchanged. Post-treatment SPECT/CT or PET/CT (Y-90 PET/CT, utilizing internal pair production) to verify microsphere distribution.

Treatment Approaches

Lobar treatment: entire right or left lobe; standard for multifocal disease. Segmental/subsegmental treatment (radiation segmentectomy): superselective delivery to 1-2 segments containing the tumor. Enables ablative doses (>400 Gy) to small tumors. LEGACY trial: radiation segmentectomy for solitary HCC <8 cm achieved complete pathological necrosis in 89% and 3-year OS of 86%. Emerging as a curative-intent treatment for early-stage HCC. Radiation lobectomy: intentionally treating an entire lobe to cause atrophy and induce contralateral hypertrophy (analogous to portal vein embolization); allows subsequent resection.

Key Trials

DOSISPHERE-01 (2020)

Personalized dosimetry (>205 Gy to tumor) vs. standard dosimetry for HCC. Significantly improved objective response rate and overall survival. Established personalized dosimetry as the new standard.

LEGACY Study (2023)

Single-arm study of radiation segmentectomy for solitary HCC (BCLC 0/A). Objective response rate: 88%; complete pathological necrosis: 89%. Supports Y-90 as a curative-intent treatment for early-stage HCC.

SARAH and SIRveNIB Trials (2017)

Y-90 (resin) vs. sorafenib for advanced HCC. No survival advantage for Y-90 over sorafenib (but Y-90 had better quality of life and fewer adverse events). Criticism: used BSA dosimetry (not personalized); many patients did not receive optimal Y-90 doses.

EPOCH Study (2024)

Y-90 glass microspheres (personalized dosimetry) vs. sorafenib for advanced HCC with PVT. Pending/recent results expected to show benefit of personalized Y-90.

Complications

Post-radioembolization syndrome: fatigue (most common, 50-70%), nausea, abdominal pain, low-grade fever; milder than post-TACE syndrome. Radiation-induced liver disease (RILD): sinusoidal obstruction syndrome from non-tumoral liver irradiation; presents as jaundice, ascites, hepatomegaly 4-8 weeks post-treatment; potentially fatal. Radiation pneumonitis: from excessive lung shunting; prevented by LSF assessment. GI ulceration: from non-target deposition in gastric/duodenal arteries; prevented by proper mapping and coil embolization. Biliary complications: radiation cholecystitis, biliary stricture. Lymphopenia: common and often persistent; clinical significance unclear.

<image>Illustration of the Y-90 radioembolization workflow. Five sequential panels: (1) Mapping angiogram showing celiac and hepatic artery anatomy with coil embolization of the gastroduodenal and right gastric arteries; (2) MAA injection through the microcatheter at the planned treatment position; (3) SPECT/CT MAA scan showing tumor uptake, normal liver background, and lung shunt fraction calculation; (4) Y-90 microsphere delivery through the microcatheter with a magnified inset showing microspheres lodged in tumor arterioles emitting beta radiation; (5) Post-treatment Y-90 PET/CT confirming microsphere deposition in the tumor. Each panel is labeled with the corresponding time point in the treatment timeline.</image>

<image>Comparison of BSA dosimetry versus personalized partition model dosimetry for Y-90. Two panels: Left panel shows BSA method with a simple formula and uniform dose delivery, resulting in suboptimal tumor dose in some cases. Right panel shows partition model with tumor, normal liver, and lung compartments receiving individualized doses, with the DOSISPHERE target of >205 Gy to the tumor highlighted. A bar graph below compares objective response rates between BSA (36%) and personalized dosimetry (71%) from the DOSISPHERE-01 trial. The tumor-to-normal-liver uptake ratio (T/N ratio) is illustrated on the MAA SPECT/CT.</image>

<image>Illustration of radiation segmentectomy versus lobar radioembolization. Two panels: Left panel shows lobar Y-90 treatment with microspheres distributed throughout the right lobe, delivering moderate dose to both tumor and normal parenchyma. Right panel shows radiation segmentectomy with superselective catheter position and ablative dose (>400 Gy) concentrated in a single segment containing a small HCC tumor, with minimal dose to the remaining liver. Post-treatment MRI insets show: lobar treatment with partial tumor necrosis; segmentectomy with complete tumor necrosis and segmental atrophy. The LEGACY trial results are annotated.</image>

Key Clinical Pearls

Y-90 is NOT embolic at therapeutic doses -- it is brachytherapy; this is why it can be used safely with portal vein thrombosis (unlike TACE). Personalized dosimetry (partition model, >205 Gy to tumor) dramatically improves outcomes compared with BSA dosimetry -- DOSISPHERE changed practice. Radiation segmentectomy is emerging as a curative-intent treatment for early-stage HCC, rivaling ablation and resection. The mapping arteriogram is the most critical step -- incomplete coil embolization of non-target vessels leads to devastating GI complications. Lung shunt fraction >15-20% is a contraindication; always calculate lung dose before treatment. Y-90 causes less post-procedural morbidity than TACE (no ischemic component); fatigue is the dominant complaint. Failed SARAH/SIRveNIB trials used suboptimal dosimetry -- do not conclude Y-90 is ineffective based on these studies.

References

• Hermann AL et al. Relationship of tumor radiation-absorbed dose to survival and response in HCC treated with TARE (DOSISPHERE-01). Gastroenterology 2020
• Salem R et al. Y-90 radiation segmentectomy for treatment of solitary HCC (LEGACY). J Hepatol 2023
• Vilgrain V et al. Efficacy and safety of selective internal radiotherapy with Y-90 resin microspheres compared with sorafenib in locally advanced and inoperable HCC (SARAH). Lancet Oncol 2017
• Salem R, Thurston KG. Radioembolization with Y-90 microspheres: a state-of-the-art brachytherapy treatment. J Vasc Interv Radiol 2006

Read this lecture as Markdown