Margriet Kwint

Appendices 176 Part II Image Guided Radiotherapy The overall treatment time of concurrent chemoradiation is about 5 to 6 weeks. In radiotherapy, it is generally assumed that the anatomy of the patient is stable during this treatment course. However, during the 5-6 weeks course of lung radiotherapy several anatomical changes may occur, such as increasing/decreasing atelectasis, infiltrative changes, tumor progression or regression and pleural effusion. With the introduction of advanced image-guided systems like kilo voltage (kV) CBCT, we have the ability to visualize the tumor and organs at risk in three dimensions just before, during and/or after each fraction. These CBCT is primarily used to minimize target misalignment and setup error. In clinical practice repetitive CBCT’s during treatment have made us aware of intra thoracic anatomical changes (ITACs) during the course of a radical treatment. The aim of the study described in chapter 4 was to quantify the ITAC’s during the radiotherapy course. A total of 1500 CBCT’s of 177 patients were analyzed. Our decision support system: “the traffic-light protocol”, was retrospectively applied to all of these CBCT-scans. The traffic-light protocol has three urgency levels: red (considerable impact on dose distribution), orange (moderate impact on dose distribution) and green (negligible impact on dose distribution). In 72% of the patients ITAC’s were observed with a maximum level of red, orange and green in 12%, 36% and 24% respectively. Fourteen patients (8%) required a new planning CT- scan and an adapted treatment plan to account for anatomical changes. Types of observed ITACs were, evident tumor regression (35%), considerable tumor baseline shift (27%), changes in atelectasis (19%), tumor progression (10%), pleural effusion (6%) and infiltrative changes (3%). This decision support system is currently used in our institute by the RTT’s who analyze the CBCT’s during treatment. Although concurrent chemoradiotherapy of NSCLC patients has a curative intent, OS remains poor. To distinguish between patients with better or worse OS, prediction models are used. Current prediction models mainly use baseline characteristics to predict treatment outcomes. An important step to improve these prediction models is to incorporate longitudinal data. The use of CBCT’s during treatment resulted in an increase of available imaging data of tumor volume change during radiotherapy treatment. The aim of the study in chapter 5 was to identify subgroups of LA-NSCLC patients showing tumor volume changes during CCRT and to investigate whether the identified subgroups are associated with treatment outcomes. In this study of 394 patients, 3 different subgroups of tumor volume change during treatment