Holmium-166 Radioembolization: SPECT Imaging and Dosimetry

During the last decade, the use of holmium-166 (166Ho) microspheres as radioembolization device has been increasing throughout Europe.

Chapter 1 presented the use of 166Ho microspheres as an alternative to 90Y for radioembolization, during the scout and the treatment, describing the steps involved in the clinical workflow. 166Ho imaging possibilities, particularly SPECT, were also presented. Additionally, the 166Ho-99mTc dual isotope protocol was introduced.

Radioembolization planning and outcome is closely related to dosimetry assessment. However, dosimetry relies on adequate imaging, especially when clinical reconstructions are used. In the first part of this thesis, the technical challenges related to 166Ho SPECT imaging in the clinical setting have been presented.

Since high activities of 166Ho–labelled microspheres are used for therapeutic radioembolization, scanning patients shortly after administration can impact gamma camera performance, possibly hampering accuracy of dosimetry. Chapter 2 assessed the impact of the high-count rate on 166Ho SPECT imaging using an anthropomorphic phantom, filled with activity up to 2.7 GBq. The high-count regime did not affect image quality when visually assessed, nor the possibility of reliable dosimetry within the healthy liver. However, tumor dosimetry was hampered due to the gamma camera dead time.

Absolute activity is not the only parameter influencing 166Ho SPECT imaging. 166Ho has a complex spectrum, with the main photopeak window contaminated by scatter from 166Ho high energy gamma emissions. In a separate study, 99mTc colloid is intravenously injected, after 166Ho administration, with the purpose of imaging the healthy liver parenchyma. This requires a dedicated dual-isotope imaging protocol to separate 166Ho and 99mTc signal into individual SPECT reconstructions. When a dual-isotope protocol is used as well, the additional isotope, 99mTc, and its main photopeak window centered at 140 keV can further influence 166Ho main photopeak. To account for the scatter coming from high energies, different scatter correction factors were tested in Chapter 3, using both phantom measurements and patient data.

Chapter 4 described the acquisition protocols and the impact of different reconstruction methods on image quality. The proposed protocols have been adopted in a multi-center study performed among different centers in The Netherlands, with the purpose of harmonizing 166 Ho SPECT image quality.

The 166Ho-99mTc dual-isotope protocol can improve dosimetry by facilitating the delineation of the healthy liver parenchyma. However, 166Ho images acquired in presence of an additional isotope should be proven to be suitable to conduct 166Ho dosimetry. Chapter 5 compared the resulting dosimetry using 166Ho images acquired in presence or not of 99mTc, assessing three volumes of interest: lungs, healthy liver and tumors.

The automatic healthy liver segmentation method presented in Chapter 3 for a phantom study was applied to patient data in Chapter 6. The resulting healthy liver masks were compared to the manual segmentation performed by an experienced physician with respect to dosimetry and overlap-indices. Relation between toxicity and healthy liver dose was also assessed.

Besides healthy liver toxicity, also radiation pneumonitis can be a side-effect of radioembolization. Radioembolization occurrence and lung dose measured on post-treatment PET following 90Y radioembolization treatment was presented in Chapter 7.