Clinical Workflow for Planning and Treating Palliative Patients on Diagnostic Images

Purpose/Objective(s) Care coordination for inpatients requiring palliative radiotherapy remains a challenge. One such challenge is timely CT simulation. Previously published data demonstrate that rapid radiotherapy delivery is associated with improved patient satisfaction and reduced duration of hospitalization. The purpose of this study was to assess the feasibility of eliminating RT simulation from the current clinical workflow for palliative patients by using diagnostic images as the primary image dataset. Materials/Methods A cohort of 11 previously treated palliative patients were identified. Treatment sites evaluated included spine (4), sacrum (2), shoulder (1), lung (1), humerus (1), pelvis (1) and whole brain (1). Diagnostic CT images on a curved couch without RT immobilization were used to generate diagnostic plans. The diagnostic plans were copied and recalculated on corresponding simulation CT images, which were the clinically used datasets for planning and treatment. For the whole brain patient who only had a diagnostic MRI, a synthetic CT scan was generated with brain, muscle, adipose and bone overridden following ICRU 44. Anatomic surrogates were aligned to couch indexing notches to avoid large shift during patient setup. The quality of diagnostic imaging based treatment plans were compared against traditional CT simulation based plans by a group of physicians. Results All diagnostic plans, except the shoulder, were clinically acceptable. Minor dose distribution deviation from the previously approved clinical plans were noted, but were not clinically significant, as illustrated in the table. The average change in maximum point dose was 0.1 Gy, ranged between -0.8 Gy and 0.5 Gy. Change in PTV D95 and V95 ranged from -1.3% to 0.5%, and from 0% to 2.8%, respectively. The rejected shoulder case was secondary to a limited field of view (FOV) on the diagnostic images. However, a limited FOV may be corrected through estimation of missing tissue or more appropriate diagnostic image selection. The impact from CT calibration and anatomy change on different couches was negligible. Conclusion A clinical workflow to generate treatment plans on diagnostic imaging was developed. This workflow eliminates the need for formal CT simulation, allowing for expedited radiotherapy. Diagnostic imaging based treatment plans were deemed to be clinically acceptable after careful quality and dosimetric assessment against traditional CT simulation based treatment plans. Limitations of the new workflow were identified and iterative improvements are currently underway.