Radioembolization with yttrium-90 (Y-90) microspheres is a treatment option for primary and metastatic liver tumors. We aim to demonstrate correlation between activity measured within the patient and the administered activity and to describe the distribution within the liver.
Matthew E. Johnson, MD, Eugene Fourkal, PhD, Iavor Veltchev, PhD, Jian Qin Yu, MD, Mohan Doss, PhD, Josh E. Meyer, MD; Fox Chase Cancer Center
Background: Radioembolization with yttrium-90 (Y-90) microspheres is a treatment option for primary and metastatic liver tumors. The activity of the microspheres is provided by the manufacturer with an uncertainty of +/– 10% and is verified by the institution prior to injection. While the amount of activity injected into the patient is known, the resultant activity and distribution within the liver remain relatively unknown. We have developed a positron emission tomography (PET)/CT-based method to identify the injected activity and distribution within the liver following radioembolization. We aim to demonstrate correlation between activity measured within the patient and the administered activity and to describe the distribution within the liver.
Methods: Twelve patients were enrolled in institutional review board (IRB)-approved prospective studies and underwent Y-90 radioembolization for primary or metastatic liver tumor(s). As part of the protocols, each patient underwent a postinfusion PET/CT for quantification of activity, measuring pair production from Y-90 beta decay. Six patients also underwent pretreatment PET/CT for quantification of background noise. Contours were delineated on the postinfusion CT with a uniform 1-cm margin to account for PET resolution and motion. Structures that were delineated included liver, up to three liver tumors, liver lobe targeted (right or left), and left kidney. For this analysis, the left kidney mean activity density (Bq/mL) was used as background. Net activity density was determined by measuring liver + 1 cm mean activity density and subtracting background activity density. The resulting net density was then multiplied by the volume to obtain the total activity in the patient.
Results: Six patients underwent treatment with SIR-Spheres, and six were treated with TheraSpheres. Measured activity was +/– 10% of the injected activity in 8 of 12 patients (67%). Three patients had measured activity 20% to 30% less than injected, and one patient had activity 40% less than injected. When localization of total liver activity was examined, a median of 98% (range: 87%–100%) was found within the targeted lobe, and a median of 40% of activity (range: 7%–99%) was localized to the contoured liver tumors.
Conclusion: In the majority of radioembolization cases examined, activity measured using this PET/CT method was able to account for the amount of injected activity with an error of +/– 10%. This method should be further studied to enable evaluation of radiation dose distribution within the liver and tumors following radioembolization.