What is transarterial radioembolization?

Transarterial radioembolization (TARE), also known as selective internal radiation therapy (SIRT), is a minimally invasive locoregional treatment that delivers yttrium-90 (Y-90) microspheres directly into the hepatic arterial supply of liver tumors.^[1] Y-90 is a pure beta-emitter (mean energy 0.9367 MeV, maximum energy 2.27 MeV) with a mean tissue penetration of 2.5 mm and a maximum penetration of 11 mm, making it ideal for localized tumor irradiation while minimizing damage to surrounding normal hepatic parenchyma.^[2] The physical half-life of 64.1 hours ensures sustained radiation delivery over approximately 14 days.

The therapeutic principle exploits the same arterial hypervascularity of hepatocellular carcinoma (HCC) that underpins transarterial chemoembolization (TACE). Microspheres measuring 20-60 micrometers lodge within the tumor microvasculature, delivering radiation doses of 40-800 Gy to tumor tissue while the adjacent normal liver receives significantly lower doses due to its predominant portal venous perfusion.^[3] Unlike TACE, TARE does not rely on embolic-induced ischemia, preserving arterial patency and enabling treatment of patients with main portal vein thrombosis—a population traditionally excluded from TACE.^[4]

Contemporary practice encompasses three distinct treatment strategies: whole-liver treatment for diffuse bilobar disease, lobar treatment for unilateral disease, and radiation segmentectomy for small, localized tumors (typically <=3 cm) where superselective delivery achieves ablative doses (>190 Gy) with curative intent.^[5] The EASL and AASLD guidelines recognize TARE as a standard-of-care option for intermediate-stage HCC (BCLC B), advanced HCC with portal vein invasion (BCLC C), and as a bridge to transplantation.^[6]

Microsphere platforms: SIR-Spheres vs TheraSphere

Two distinct Y-90 microsphere platforms are commercially available, each with unique physical characteristics, dosimetry models, and regulatory indications. Understanding these differences is essential for safe and effective treatment planning.^[8]

SIR-Spheres (Sirtex Medical) are resin-based microspheres with an average diameter of 32.5 micrometers (range 20-60 um). Each vial contains approximately 40-80 million microspheres with a total activity of 3 GBq. The resin platform uses a body surface area (BSA) method for dose calculation, modified by lung shunt fraction and tumor-to-liver ratio. SIR-Spheres are indicated for the treatment of unresectable metastatic liver tumors from primary colorectal cancer with adjuvant intrahepatic chemotherapy, and are widely used off-label for HCC.^[9]

TheraSphere (Boston Scientific) are glass microspheres with a smaller average diameter of 25 micrometers (range 20-30 um) but significantly higher specific activity per sphere (2500 Bq vs 50 Bq for resin). Each vial contains 3-8 million microspheres with total activity ranging from 3-20 GBq depending on calibration. TheraSphere uses a compartmental dosimetry model based on the Medical Internal Radiation Dose (MIRD) approach, calculating absorbed dose to tumor and normal liver tissue. TheraSphere holds specific FDA pre-market approval for HCC treatment.^[10]

The practical implications are substantial. Glass microspheres require fewer particles to deliver the same activity, potentially reducing the risk of stasis-related complications. Resin microspheres distribute more diffusely due to higher particle numbers, which may be advantageous for infiltrative or multifocal disease. Dosimetry transition from BSA-based to personalized MIRD-based approaches is increasingly recommended regardless of platform, as it correlates dose with objective response and survival.^[11]

Patient selection and indications

Transarterial radioembolization is indicated across the full spectrum of BCLC staging in appropriately selected patients. For BCLC B (intermediate-stage) HCC, TARE offers an alternative to TACE, particularly when tumor burden is diffuse, portal vein thrombosis is present, or TACE refractoriness has developed.^[12] For BCLC C (advanced-stage) HCC, TARE provides locoregional control with fewer systemic side effects than sorafenib, as demonstrated in the SARAH and SIRveNIB randomized trials.

Key indications include: unresectable HCC (BCLC B/C), HCC with main or branch portal vein tumor thrombosis (PVTT), intrahepatic cholangiocarcinoma (ICC), chemorefractory colorectal liver metastases, neuroendocrine tumor liver metastases, and bridge-to-transplantation for patients awaiting liver allocation.^[13] The LEGACY study demonstrated that radiation segmentectomy with TheraSphere achieved complete response in 88.2% of solitary HCCs <=3 cm, with median overall survival of 26.2 months—outcomes comparable to surgical resection and thermal ablation.^[14]

Patient selection requires multidisciplinary evaluation incorporating hepatology, surgical oncology, transplant surgery, medical oncology, and interventional radiology. Performance status (ECOG 0-1), hepatic functional reserve (Child-Pugh A-B7), and adequate lung function (FEV1 >1 L) are prerequisites. The ALBI grade and MELD score provide objective risk stratification. Tumor burden, extrahepatic disease status, and prior hepatic radiation must be evaluated.^[15]

Selection criteria for optimal outcomes

Ideal candidates exhibit preserved hepatic function (Child-Pugh A, ALBI grade 1-2), ECOG performance status 0-1, tumor burden <50% of liver volume, and absence of significant extrahepatic disease. Patients with main PVTT may be treated if Child-Pugh A and adequate collateral portal flow is confirmed. Bilirubin >2 mg/dL is a relative contraindication for whole-liver or lobar treatment but may be acceptable for segmental approaches. Prior extensive liver external beam radiation (>30 Gy) requires careful dosimetric evaluation.^[16]

MAA planning angiogram and lung shunt assessment

A mandatory 99mTc-macroaggregated albumin (MAA) planning angiogram must be performed 1-2 weeks before Y-90 delivery.^[17] This procedure serves three critical functions: mapping hepatic arterial anatomy for treatment planning, quantifying hepatopulmonary shunt fraction (LSF), and identifying extrahepatic deposition that could cause nontarget radiation injury to the gastrointestinal tract, gallbladder, or lungs.

The planning procedure follows selective hepatic arteriography with catheter or microcatheter placement in the intended treatment position. 99mTc-MAA (4-5 mCi, 148-185 MBq) is injected through the catheter, followed by planar and SPECT/CT imaging. The SPECT/CT component is essential for accurate LSF quantification and three-dimensional localization of extrahepatic deposition.^[18] Cone-beam CT during the planning angiogram further refines vascular mapping and identifies variant anatomy, parasitic feeders, and dangerous anastomoses.

The lung shunt fraction (LSF) represents the percentage of injected activity that reaches the pulmonary circulation via tumor-related arteriovenous shunting. Safety thresholds vary by platform: for resin microspheres, LSF >20% is generally contraindicated, while for glass microspheres, LSF >10% requires dose reduction or treatment modification.^[19] The lung dose limit is 30 Gy per treatment and 50 Gy cumulative. Extrahepatic deposition to the stomach, duodenum, or pancreas must be eliminated through coil embolization of feeder vessels (gastroduodenal artery, right gastric artery, falciform artery) or catheter repositioning before Y-90 delivery.

Dosimetry models and dose calculation

Contemporary TARE practice has transitioned from empirical dosing to personalized dosimetry based on the MIRD formalism.^[21] Three primary dosimetry approaches are employed:

1. Body Surface Area (BSA) method (traditionally used with resin microspheres): Activity (GBq) = (BSA - 0.2) + (tumor volume / total liver volume). This method is simple but does not account for individual tumor biology or lung shunt, leading to variable tumor doses.

2. Compartmental MIRD method (standard for glass microspheres): Calculates absorbed dose to tumor (Gy) and normal liver (Gy) based on injected activity, liver mass, and tumor-to-liver uptake ratio. The partition model assumes uniform distribution within each compartment. Target tumor dose is typically 120-150 Gy for palliative intent and >190 Gy for radiation segmentectomy.^[22]

3. 3D voxel-based dosimetry (emerging standard): Uses post-treatment Y-90 PET/CT or bremsstrahlung SPECT/CT to reconstruct actual microsphere distribution, enabling precise dose-volume histograms for tumor and organs at risk. This approach correlates strongly with histopathologic response and is increasingly used for treatment optimization.^[23]

The DOSISPHERE-01 trial demonstrated that personalized dosimetry targeting >205 Gy to tumor significantly improved objective response rate (71% vs 36%) and median overall survival (26.6 vs 10.7 months) compared to standard BSA dosing, establishing personalized dosimetry as the new standard of care.^[24]

Treatment delivery technique

Y-90 delivery is performed in a radiation-protected angiography suite with appropriate shielding, monitoring, and waste management protocols. Arterial access is obtained via the common femoral artery or radial artery under local anesthesia and moderate sedation. A 4-5 French vascular sheath is placed, and the treatment catheter is advanced to the pre-planned position confirmed during MAA mapping.^[25]

The Y-90 delivery system consists of a vial containing the microspheres in sterile water, connected to the patient via a dedicated administration set with inline filters. The microsphere suspension is delivered slowly over 5-10 minutes under continuous fluoroscopic monitoring. The infusion must be slow and intermittent to prevent microsphere aggregation and nontarget embolization. The catheter is flushed with sterile saline between aliquots to ensure complete delivery.^[26]

Radiation safety protocols are mandatory. All personnel must wear dosimeters. The patient is monitored for vital signs throughout the procedure. Post-delivery, the administration set, catheters, and any residual material are handled as radioactive waste per institutional radiation safety protocols. The patient may be discharged same-day or after overnight observation depending on institutional practice and patient status.

Post-treatment imaging

Bremsstrahlung SPECT/CT or Y-90 PET/CT is performed within 24 hours to confirm intrahepatic deposition, exclude nontarget delivery, and enable 3D dosimetry. Y-90 PET/CT offers superior spatial resolution and quantitative accuracy compared to bremsstrahlung imaging, allowing voxel-based dose reconstruction. The imaging confirms treatment coverage and provides baseline for response assessment at 4-6 weeks.^[27]

Radiation segmentectomy for small HCC

Radiation segmentectomy represents a paradigm shift in TARE, transforming it from a palliative to a potentially curative modality for small, localized HCC.^[28] By superselectively catheterizing a single segmental or subsegmental hepatic artery branch, operators deliver ablative radiation doses (>190-400 Gy) to tumors <=3 cm while limiting normal liver exposure. The LEGACY study reported complete response in 88.2% of solitary HCCs <=3 cm treated with radiation segmentectomy, with median overall survival of 26.2 months.^[14]

Patient selection for radiation segmentectomy requires solitary or oligofocal disease (<=3 lesions, each <=3 cm), preserved hepatic function (Child-Pugh A), and technically feasible superselective catheterization. The procedure demands advanced microcatheter skills and CBCT confirmation of target perfusion territory. Dose escalation beyond 400 Gy may be tolerated for very small lesions (<2 cm) given the limited normal liver volume at risk.^[29]

Comparative studies suggest radiation segmentectomy achieves outcomes comparable to thermal ablation (RFA, MWA) and surgical resection for small HCCs, with the added advantage of treating lesions adjacent to major vessels (where heat sink limits thermal ablation) and near the gallbladder or diaphragm (where thermal injury risks are higher).^[30]

Expected outcomes and response assessment

Response assessment follows mRECIST criteria on multiphasic CT or MRI at 4-6 weeks post-TARE, with subsequent imaging every 2-3 months.^[31] Unlike TACE, where lipiodol retention may confound CT interpretation, TARE produces no radiopaque artifact, enabling clear evaluation of residual arterial enhancement. MRI with hepatobiliary-specific contrast (gadoxetic acid) is preferred for HCC assessment due to superior detection of subtle residual tumor.

For HCC (BCLC B/C), objective response rates (CR + PR by mRECIST) range from 40-60% following lobar or whole-liver TARE, with median overall survival of 12-18 months.^[32] The SARAH trial (n=459) demonstrated non-inferior overall survival of TARE (10.0 months) versus sorafenib (9.9 months) for advanced HCC, with significantly improved quality of life and fewer adverse events. The SIRveNIB trial (n=360) confirmed these findings in the Asia-Pacific population.^[7]

For radiation segmentectomy, complete response rates of 60-90% are reported for small HCCs, with 3-year overall survival of 60-80%.^[14] Downstaging to resection or transplantation occurs in 10-20% of appropriately selected patients. For colorectal liver metastases, TARE improves progression-free survival when added to systemic chemotherapy (first-line FOLFOX), though overall survival benefit remains modest.^[33]

Beyond anatomic response, functional imaging with 18F-FDG PET/CT may predict outcomes earlier than structural imaging. A decline in tumor SUVmax >50% at 4-6 weeks correlates with improved overall survival. Serum AFP response similarly provides prognostic information complementary to imaging.^[34]

Complications and REILD management

TARE is generally well tolerated, with a distinct adverse event profile compared to TACE. The most significant complication is radiation-induced liver disease (REILD), a clinical syndrome of jaundice, ascites, and hepatomegaly occurring 4-8 weeks post-treatment in 5-10% of patients.^[35] REILD results from radiation injury to hepatic sinusoidal endothelial cells and is more common in patients with pretreatment bilirubin >2 mg/dL, whole-liver treatment, or prior hepatic radiation. Prophylactic ursodeoxycholic acid and corticosteroids may reduce incidence, though evidence is limited.

Radiation gastritis and duodenitis occur when nontarget deposition delivers >10-15 Gy to the gastric or duodenal mucosa. Symptoms include nausea, epigastric pain, and ulceration, typically manifesting 1-4 weeks post-treatment. Proton pump inhibitors (PPIs) should be initiated 1 week before treatment and continued for 4-8 weeks. Severe cases may require endoscopic management.^[36]

Radiation pneumonitis is rare but potentially fatal when lung dose exceeds 30 Gy per treatment or 50 Gy cumulative. Strict adherence to LSF thresholds and lung dose limits prevents this complication. Other complications include biliary necrosis/stricture (2-5%), portal vein thrombosis (rare), cholecystitis (if cystic artery treated), and post-radioembolization syndrome (fatigue, nausea, low-grade fever) occurring in 20-50% of patients and typically resolving within 1-2 weeks.^[37]

Follow-up imaging and surveillance

Multiphasic CT or MRI is obtained at 4-6 weeks post-TARE to assess initial response, followed by imaging every 2-3 months for the first year and every 3-6 months thereafter.^[39] MRI with hepatobiliary-specific contrast (Eovist/Primovist) is preferred for HCC due to superior detection of residual viable tumor and delayed phase washout characteristics. Y-90-induced changes in the treated parenchyma—diffuse T2 hyperintensity, perfusion alterations, and delayed enhancement—should not be mistaken for tumor progression.

Liver function tests (bilirubin, albumin, INR, AST/ALT) are monitored at 4-6 weeks to detect REILD. A rising bilirubin with stable or improving imaging should raise suspicion for REILD rather than tumor progression. AFP and other tumor markers are measured at each follow-up visit.^[40]

For patients treated with curative intent (radiation segmentectomy), surveillance mirrors that of resection or ablation: every 3 months for 2 years, then every 6 months. For palliative treatment, imaging frequency is guided by clinical status and treatment plan. Transition to systemic therapy (atezolizumab/bevacizumab, lenvatinib) is considered when mRECIST progression is confirmed or REILD precludes further locoregional therapy.^[41]

Contraindications and safety limits

Absolute contraindications to transarterial radioembolization include: lung shunt fraction >20% (resin) or >10% (glass) that cannot be reduced by dose modification, pretreatment bilirubin >2 mg/dL for whole-liver or lobar treatment (relative for segmental), extrahepatic deposition that cannot be prevented by coil embolization or catheter repositioning, prior extensive liver external beam radiation (>30 Gy), active uncontrolled infection, and uncorrectable coagulopathy.^[42]

Relative contraindications include: significant extrahepatic disease (though TARE may still control intrahepatic burden), ECOG performance status >2, Child-Pugh C cirrhosis, significant portal hypertension with variceal bleeding risk, and pregnancy. For patients with main PVTT, TARE is feasible if Child-Pugh A and adequate collateral flow is present—this represents a key advantage over TACE.^[43]

Safety dose limits are: tumor dose >120-150 Gy (palliative), >190-400 Gy (segmentectomy), normal liver dose <40-50 Gy, lung dose <30 Gy per treatment and <50 Gy cumulative, and stomach/duodenum dose <10-15 Gy. These limits guide treatment planning and dose modification when anatomic constraints preclude ideal catheter positioning.^[44]

Conclusion

Transarterial radioembolization has evolved from a palliative option to a precision locoregional therapy with curative potential for select patients with hepatocellular carcinoma. The transition from empirical BSA-based dosing to personalized MIRD and 3D voxel-based dosimetry—exemplified by the DOSISPHERE-01 trial—has fundamentally improved outcomes, establishing dose-response relationships that guide contemporary practice.

The unique ability to treat HCC with main portal vein thrombosis, the favorable safety profile compared to systemic therapy, and the curative outcomes of radiation segmentectomy for small tumors position TARE as an essential modality in the interventional oncology armamentarium. For radiologists, radiographers, and hospital administrators, implementing rigorous MAA planning protocols, personalized dosimetry workflows, and standardized follow-up surveillance ensures optimal patient selection, treatment delivery, and outcome measurement across the service line.