Role of multimodality imaging in radiotherapy planning

Accurate tumor delineation is crucial to achieve successful radiotherapy (RT). While CTs are clinically used to plan 3D–RT, multimodality imaging (MMI) promises to aid accurate delineation of the tumor. This thesis examined delineated MMI volumes and studied their role in RT planning. A prostate modeling study evaluated the dose reductions in conformal and intensity‐modulated radiotherapy (CRT and IMRT) due to interfractional gross tumor volume (GTV) motion and variability in delineation. In both CRT and IMRT, a 10 mm margin was adequate to account for interfractional GTV movement. However, 10 mm (CRT) and 5 mm (IMRT) GTV misdelineation produced unacceptable dose distributions. IMRT plans were more affected by delineation errors when compared to CRT plans due to their steep dose gradients. Methods were developed to use MRI only in RT planning. Water and bone electron density values were assigned to the image and a low‐distortion MR‐sequence was used. MR‐delineated prostate volumes were in general smaller than those on CT by 10%. Patient data were separated into those that were affected by organ motion from those that were not. An independent‐organ registration technique was developed to relate CT‐ and MR‐delineated volumes that were affected by organ motion between the scans. The dosimetric study showed that unacceptable CRT was planned using planning target volume delineated using CT when volumes were assumed to be true. However, the dosimetric benefit was small as a 1–2 mm marginal increase on would deliver adequate dose to both PTVs and insignificant dose reductions were observed in rectum and bladder when plans were compared to plans. The dosimetric effect of in neuroblastoma CRT and IMRT was studied. Reduction in tumor control probability due to incorrectly delineated neuroblastoma was observed and a new treatment was planned using with acceptable doses to OARs the organs‐at‐risk. IMRT delivered more conformal dose distributions to complex‐shaped PTVs compared to CRT. Methods required to add currently acquired diagnostic to thoracic lymphoma RT planning were developed. The addition confirmed the inferior disease extension in 60% of the cases studied and in 20% of cases, presented a large inferior difference. The addition of FDG–PET would affect the field sizes and the extension of lead blocking. A theoretical CRT study showed that greater than 50% decrease in lung dose was achievable with when compared to the original plan. This thesis highlighted not only the importance of using MMI in RT planning but developed methods to cope with the additional information. Though not all the imaging modalities added a significant dosimetric benefit, they added new information regarding GTV determination and decreased the subjectivity of GTV delineation with minimal technical difficulty, which will benefit current RT planning.