End-to-end test for fractionated online adaptive MR-guided radiotherapy using a deformable anthropomorphic pelvis phantom

Objective. In MR-guided radiotherapy (MRgRT) for prostate cancer treatments inter-fractional anatomy changes such as bladder and rectum fillings maybe corrected by an online adaption of the treatment plan. To clinically implement such complex treatment procedures, however, specific end-to-end tests are required that are able to validate the overall accuracy of all treatment steps from pretreatment imaging to dose delivery. Approach. In this study, an end-to-end test of a fractionated and online adapted MRgRT prostate irradiation was performed using the so-called ADAM-PETer phantom. The phantom was adapted to perform 3D polymer gel (PG) dosimetry in the prostate and rectum. Furthermore, thermoluminescence detectors (TLDs) were placed at the center and on the surface of the prostate for additional dose measurements as well as for an external dose renormalization of the PG. For the end-to-end test, a total of five online adapted irradiations were applied in sequence with different bladder and rectum fillings, respectively. Main results. A good agreement of measured and planned dose was found represented by high gamma-index passing rates (3%/3mm criterion) of the PG evaluation of 98.9% in the prostate and 93.7% in the rectum. TLDs used for PG renormalization at the center of the prostate showed a deviation of - 2.3%. Significance. The presented end-to-end test, which allows for 3D dose verification in the prostate and rectum, demonstrates the feasibility and accuracy of fractionated and online-adapted prostate irradiations in presence of inter-fractional anatomy changes. Such tests are of high clinical importance for the commissioning of new image-guided treatment procedures such as online adaptive MRgRT.