Clinical dosimetry in molecular radiotherapy
Team : Multi-scale dosimetry for radiotherapy optimization
Molecular Radiotherapy (MRT) is a systemic radiotherapy, where the radiopharmaceutical binds specifically to tumours to selectively destroy cancer targets while preserving healthy organs. Lutathera® (177Lu-DOTATATE) is a radiopharmaceutical recently approved by the FDA/EMA for the treatment of gastroenteropancreatic neuroendocrine tumours (GEP-NETs). In clinical practice, patients receive a fixed activity of Lutathera®, 4 cycles of 7.4 GBq, assuming that the pharmacokinetics of the radiopharmaceutical are the same between patients. Patient-specific dosimetry enables a major paradigm shift in TMR delivery from a “one size fits all” approach to true personalised medicine where the activity delivered is specifically assessed on the basis of the radiation delivered to each patient. This usually involves determining the spatial distribution of the radiopharmaceutical in the organs by imaging at different times (quantitative imaging), estimating the total number of radioactive decays by integrating the activity over time (pharmacokinetic assessment) and calculating the absorbed dose from the physical characteristics of the radionuclide and energy transport in the patient’s tissue. Currently, there are no standardised procedures for performing clinical dosimetry. Furthermore, the evaluation of uncertainties associated with the dosimetry procedure is not trivial. The DosiTest project was initiated to evaluate the uncertainties associated with each step of the clinical dosimetry workflow, via a multicentre inter-comparison based on Monte Carlo (MC) modelling. The first phase of the thesis consisted in comparing dosimetric analyses performed by different centres using the same software and protocol on the same patient dataset in the framework of the IAEA-CRP E23005 project in order to assess the accuracy of clinical dosimetry. To our knowledge, this is the first time that a multi-centre dosimetric comparison of a single set of clinical patient data has been undertaken using the same protocol and software by multiple centres worldwide. It highlighted the critical need for checkpoints and common sense checks to eliminate significant disparities between results and to distinguish malpractice from acceptable inter-operator variability. An important outcome of this work was the lack of quality assurance in clinical nuclear medicine dosimetry and the need to develop quality control procedures. As dosimetry gains popularity in nuclear medicine, best practice must be adopted to ensure reliability, traceability and reproducibility of results. This also highlights the need for sufficient training after the acquisition of relatively new software packages, beyond a few days. This is clearly insufficient in the context of an emerging field where professional experience is often lacking. Secondly, the study of the accuracy of clinical dosimetry requires the generation of test data sets, in order to define the ground truth against which clinical dosimetry procedures can be compared. The second section of the thesis deals with the simulation of three-dimensional SPECT imaging by implementing the self-contouring detector motion in the GATE Monte Carlo toolbox. After validation of the SPECT/CT projections on anthropomorphic models, a series of realistic images of clinical patients was generated. The last part of the thesis established the proof of concept of the DosiTest project, using a virtual (simulated) SPECT/CT data set at different times, with different gamma cameras, allowing to compare different dosimetric techniques and to evaluate the clinical feasibility of the project in some nuclear medicine departments.
Thesis defended on 31 march 2022