This trial involves patients with hepatocellular carcinoma (HCC) diagnosis treated with transarterial radioembolization (TARE) with resin microspheres loaded with 90Y. Patients will be divided in two groups based on prescription method to calculate the therapeutic radionuclide activity to be injected. In arm A, standard dosimetric approach such as Body Surface Area (BSA) method and Medical Internal Radiation Dosimetry (MIRD) monocompartmental method will be used. In arm B, novel voxel-based dosimetry, based on pre-treatment simulation with 99m-Technetium (99mTc)-Macro Aggregated Albumin (MAA) injection and SPECT/CT image acquisition, will be used. The primary outcome will be the overall survival of patients included in arm A and arm B. Secondary outcomes will be adverse events, tumor response, biomarkers assessed from blood samples prior and after the treatment and voxel-based dosimetry obtained from post-treatment PET/CT images acquisitions.

The hepatocellular carcinoma (HCC) is the most common type of primary liver cancer. For a large number of patients surgery is not feasible, due to disease multifocality or liver impairment. Several techniques have been proposed in the last years to treat inoperable HCC, including transarterial radioembolization (TARE), also known as selective internal radiotherapy (SIRT). The TARE is an interventional procedure in which micron-sized embolic particles (resin or glass microspheres) loaded with beta-emitter radionuclides of 90Y are directly injected in the hepatic arteries. This allows to treat the tumor, which receives most of its blood supply from these arteries, and to spare the liver parenchyma, which is supplied mainly by the portal vein. Microparticles are directly released to the target and remain trapped in it, locally delivering a curative amount of radiation dose. In addition to the dose delivered to the tumor and to the liver parenchyma, which have been showed to be related to the treatment efficacy and to the occurrence of liver toxicity, several studies highlighted the role of biomarkers such as the alpha-fetoprotein or the PIVKA-II to describe the tumor progression or aggressiveness.

The procedure can be simulated with the injection of 99mTc-labeled macroaggregated albumin (MAA), through the acquisition of a SPECT/CT, and verified after the treatment with bremsstrahlung SPECT imaging or PET imaging.

To assess the amount of radionuclide to be injected for a curative purpose (namely, the activity at the injection time), several methods have been proposed. Among them, the Body Surface Area (BSA) method allows to estimate the activity based on patient's height and weight and liver/tumor volumes, while the Medical Internal Radiation Dosimetry (MIRD) monocompartmental method allows to estimate the activity based on the desired dose delivered to the target or liver, assuming that all the activity is deposited inside these volumes. Both methods currently represent the standard dosimetric approach.

In addition, in recent years, voxel based dosimetry has been proposed as a novel dosimetric approach. In voxel based dosimetry, the absorbed dose distribution can be calculated from the SPECT/CT acquisition obtained in the simulation procedure prior to the treatment. This allows to take into account the inhomogeneity of the radionuclide distribution among the liver and the target volumes. Voxel based dosimetry can be used to prescribe the activity at injection time that would allow to deliver the desired absorbed dose to the target without exceeding the radiation dose to the healthy tissues.

In this randomized trial, patients with hepatocellular carcinoma (HCC) diagnosis are equally divided into arm A (activity prescription based on standard dosimetry approach) and arm B (activity prescription based on novel voxel-based dosimetry approach). The primary goal of the study is to show the superiority of the novel dosimetry approach in terms of patients' overall survival. As secondary goal, the following data will be recorded: adverse events, tumor response, biomarkers assessed from blood samples prior and after the treatment and voxel-based dosimetry obtained from post-treatment PET/CT images acquisitions.

Inclusion Criteria:

• hepatocellular carcinoma diagnosis (BCLC stage B or C)
• appropriate liver function (Child-Pugh A or B<8)
• life expectancy > 3 months
• bilirubin < 2.0 ng/ml
• multidisciplinary team consensus to the treatment
• multiphasic CT within 1 month prior to the treatment
• written informed consent

Exclusion Criteria:

• lung shunt fraction > 20%
• Child-Pugh score C
• BCLC stage D
• liver impairment
• relevant ascites
• hepatic encephalopathy
• vascular abnormality not allowing proper catheterization
• significant distant metastasis