Quantitative imaging of Pb-212

Introduction: Pb-212 (t_1/2=10.6 hours) is a beta emitter that acts as an in vivo generator of alpha particles via its short-lived progenies Bi-212 (t_1/2=60.6 min) and Po-212 (t_1/2=0.3 µs). Methods exist to produce Pb-212 at an industrial scale and chelate it to targeting agents, making it a promising radionuclide for targeted alpha therapy. In preparation for an upcoming clinical trial, the feasibility of accurate, quantitative SPECT/CT imaging of Pb-212 has been investigated using phantoms. Quantitative imaging of Pb-212 may be used to facilitate the development of methods for individualised patient dosimetric estimation.

Methods: Four 30 minute imaging protocols were examined on a Siemens Symbia Intevo Bold SPECT/CT, using two collimators, high energy (HE) and medium energy low penetration (ME), and two energy windows, 40% centered at 79 keV and 20% at 239 keV, with 20% and 5% dual scatter windows respectively. Seven reconstructions were performed, (iterations x subsets) 10x1, 15x1, 30x1, 30x2, 30x3, 30x4, 30x30, and the images were studied both without a filter and with a 12 mm Gaussian filter. Calibration factors (CFs) were found for each protocol by repeated imaging of a 6283 ml uniform, cylindrical phantom filled with 0.5-8.5 MBq Pb-212. The NEMA IEC PET Body Phantom was imaged with the same protocols to find mean and maximum recovery coefficients (RC_mean and RC_max respectively) and their fractional uncertainties. The imaging was repeated eight times for each imaging protocol with the spheres containing 17.2-104.2 kBq/ml Pb-212 and no background activity. Spherical VOIs of the physical phantom volume were placed using CT images.

Results: For all imaging protocols examined, the CFs (in cps/kBq) were found to be approximately constant when the total imaged activity exceeded 1 MBq, below which they sharply increased. The CFs were similar between reconstructions and collimators, but significantly larger for the 79 keV than the 239 keV window. The ME 79 keV protocol gave the largest count rate, with CFs ranging from 0.38±0.06 cps/kBq to 0.53±0.06 cps/kBq for the different reconstructions. In comparison, the CFs ranged from 0.28±0.03 cps/kBq to 0.35±0.03 cps/kBq for the HE 79 keV protocol, from 0.05±0.03 cps/kBq to 0.08±0.01 cps/kBq for ME 239 keV and from 0.06±0.02 cps/kBq to 0.08±0.02 cps/kBq for HE 239 keV. The activity quantification of small volumes gave similar fractional uncertainties for the all protocols except HE 239 keV, which gave larger fractional uncertainties. For ME 79 keV, HE 79 keV, and ME 239 keV activities in volumes of 2.6 ml were quantified with uncertainties below 20% with the 30x4 reconstruction. The quantitative ability was comparable for filtered and unfiltered images, but filtered images were superior for visual interpretation. More spheres were visible with the 79 keV window and the HE 79 keV protocol gave the visually best images with a low activity concentration.

Conclusions: In this setting, quantitative SPECT imaging of Pb-212 is possible with activities larger than 1 MBq for all the investigated protocols. Our results indicate that the activity in volumes as small as 2.6 ml can be reproduced with an uncertainty subceeding 20%. While there are only minor differences between many of the protocols, the 79 keV window appears to be the best choice for quantitative imaging, and the HE 79 keV protocol seems to produce the best visual results.