90Y PET-based dosimetry after selective internal radiotherapy treatments.

Objectives: The decay of 90Y has a minor branch to the O+ first excited state of 89Zr, the de-excitation of which to the fundamental state is followed by a β+–β- emission that has been used recently for biodistribution assessment after selective internal radiotherapy (SIRT) treatments. The purpose of the present study is to demonstrate the feasibility of 90Y PET imaging for dose assessment after radioembolization with 90Y microspheres.

Methods: Activity quantification was validated through preliminary phantom studies using a cylindrical body phantom composed of six inserts of different volumes filled with a calibrated amount of 90Y microspheres. A GE Discovery ST PET/CT scanner provided with bismuth germinate (BGO) crystals was used for image acquisition. Images were reconstructed with an ordered subset expectation–maximization method. The effect of object size and the effect of the number of iterations on dose evaluation and volume recovery were investigated. Microsphere dose distribution was then evaluated on one patient (one lesion) who underwent liver SIRT treatment. Dose calculations were made with a MATLAB-based code developed in our department. Dedicated Monte Carlo calculations were executed to evaluate dose S-values for the 90Y source. The activity distribution derived from 90Y PET acquisitions was convolved with the voxel S-values to obtain a three-dimensional absorbed dose distribution and dose–volume histograms.

Results: Dosimetry studies carried out on the body phantom with ordered subset expectation–maximization algorithm, three iterations, provided an accuracy of 7.62% in determining the absorbed dose in the largest insert. The dose difference increases as the insert size reduces. Preliminary results on a patient provided a high-resolution absorbed dose distribution map. An average dose of 139.3 Gy was evaluated for the tumor area, with a maximum dose as high as 237.9 Gy. The absorbed dose to the healthy liver was below the tolerance dose of 35 Gy (33.8 Gy). A clear correlation between absorbed dose and tumor response was observed at 18F-fluorodeoxyglucose PET acquired 6 months after treatment.

Conclusion: According to our experience, 90Y PET is a promising and reliable technique for microsphere dose assessment and might pave the way for a patient-specific PET-based dosimetry after liver SIRT treatments.