Comparison of the 90Y-labelled glass microspheres liver radioembolisation dosimetry with the estimated dosimetry obtained from pre-treatment 99mTc-MAA SPECT images reconstructed with and without attenuation correction

Abstract

ABSTRACT90Y-labelled glass microspheres liver radioembolisation is an effective treatment for unresectable hepatic tumours. Its treatment planning is based on 99mTc-macro-aggregated albumin (MAA) single photon emission computed tomography (SPECT) imaging. However, photons interaction with human tissues may significantly compromise the accuracy of radioactivity quantification in nuclear medicine. The aim of this work is to evaluate the impact of a patient-specific attenuation correction on pre-treatment 99mTc-MAA SPECT images and, consequently, on the estimation of radioembolisation dosimetry. The methodology developed for attenuation correction was firstly validated on phantom models. Following this, a data set of eight patients was retrospectively selected from our database and the pre-treatment images were reconstructed with and without attenuation correction. Then, the liver dosimetry maps were estimated and compared with the liver dosimetry maps estimated from the post-treatment 90Y positron emission tomography images. Results show that the attenuation corrected phantom images were greatly improved, originating higher contrast, signal-to-noise ratio and contrast-to-noise ratio comparatively to non-attenuation corrected images. Attenuation corrected clinical pre-treatment images significantly increase the agreement, measured with the intraclass correlation coefficient, between predictive and post-treatment dosimetry maps comparatively to the non-attenuation corrected ones (Wilcoxon test, p = 0.008). In conclusion, a statistically significant increase in the agreement was achieved; however, poor correlation is still observed at the single-voxel level in some patients. Thus, future research should be addressed to investigate the reasons of this and manage to optimise the agreement between the predictive and post-treatment dosimetry.