• Improved prediction of haematological toxicity during 177Lu Lu-DOTA-TATE therapy by optimising serial SPECT-CT red marrow dosimetry

      Tipping, Jill; Page, Emma; Calvert, Nicholas; Hamilton, David; Cullen, D.; Price, E.; Pells, S.; Needham, G.; Manoharan, Prakash; The Christie NHS Foundation Trust, Manchester (2020)
      Aim/Introduction: This research compared red marrow absorbed dose estimations following Peptide Receptor Radionuclide Therapy (3D Image Based (IB) versus Blood Based (BB) methods), and dose correlations with haematological toxicity. Materials and Methods: Data were acquired during cycle 1 of [177Lu]Lu-DOTA-TATE therapy for a cohort of PRRT patients. Red marrow dosimetry was performed using BB methods and IB methods using serial SPECT/CT methods, with dose uncertainties estimated. The effect on total dose (due to self & cross-irradiation) was assessed for two different cross-irradiation models (simplistic versus realistic). The strength and significance of the correlation between dose and post therapy platelet toxicity were assessed for each dosimetry method. The effect of exposure to prior treatments on dose toxicity correlations was also assessed. Results: Early results indicate that IB methods delivered significantly higher red marrow absorbed doses than BB methods (mean increase of a factor of 3 when cross-irradiation was included). Choice of cross-irradiation model had a significant impact on total dose magnitude for both BB and IB techniques. Dose response analysis of these initial results indicate that the simplistic cross-irradiation model gave the strongest and most significant dose - toxicity correlations for both IB and BB methods. The most realistic activity distribution model produced significant correlations only with IB methods for therapy naïve patients. Conclusion: Red marrow absorbed doses measured from serial SPECT were significantly higher than those from serial blood measurements. IB absorbed doses produced stronger and more significant correlations for therapy naïve patients than BB absorbed doses
    • Towards a comprehensive validation of a Monte Carlo simulation for clinical SPECT acquisition protocol optimisation and image correction

      Pells, S.; Cullen, D. M.; Robinson, Andrew P; Pietras, B.; Calvert, Nicholas; Deidda, D.; Fenwick, A.; Ferreira, K.; Hamilton, David; Heetun, W.; et al. (2020)
      Aim/Introduction: Monte Carlo (MC) simulations are a valuable tool for the optimisation of acquisition protocols and image corrections in Single-Photon Emission Computed Tomography (SPECT). Simulations allow direct gamma tracking and show all interactions that occur during acquisition. They offer an unparalleled insight into processes that are not directly experimentally observable such as scatter and attenuation within the acquisition. As the need for quantitative SPECT becomes increasingly important for dosimetry calculations, accurate image corrections are essential, and their methodologies are increasingly based on results from MC. Whilst MC is often used to provide a ‘ground-truth’, this is only the case if the simulation is fully validated against experimental data. This is a vital part of confirming the performance of new scanners. This work aims to establish a validation protocol for MC SPECT simulations. Materials and Methods: A full MC simulation of the triple-head SPECT system installed at the National Physical Laboratory has been developed in the GATE (Geant4 Application for Tomographic Emission) [1] tool-kit. The simulation extensively models the components of each detector head and a range of collimators according to technical specifications provided. The simulation can run in both dual- and triple-head acquisition modes. Validation data has been collected in a range of geometries, including commercial and 3D-printed phantoms, for multiple therapeutic and diagnostic isotopes. Results: Experimental and simulated data have been compared for both the dualand triple-head acquisitions, providing validation for all key experimental observables for a range of therapeutic and diagnostic isotopes. Conclusion: The validation procedure has demonstrated that the MC simulation accurately reproduces SPECT data. The validated simulation will be used to optimise clinical imaging protocols and image corrections, including for novel radioisotopes such as Tb- 155 and Tb-161 where low availability of isotopes limits experimental studies.