Margriet Kwint

General introduction and outline of thesis 1 11 66.3%) (12). Therefore, adjuvant immunotherapy after CCRT, in patients without tumor progression, is standard of care in the Netherlands since 2019. The studies described in this thesis have included patients who were treated between 2008 and 2017, when CCRT alone was standard of care for LA-NSCLC. NSCLC is a heterogeneous disease and together with the increase of treatment options such as chemotherapy, molecular targeted agents, immunotherapy, optimized radiotherapy schemes/techniques, and new surgery techniques, it is important to select the best treatment (or combinations) for each individual patient. Moreover, the emergence of novel parameters such as genomics, imaging modalities and new biomarkers calls for more innovative models to depict the best treatment for each patients, while taking into account several interdependencies between risk markers. Current prediction models use baseline characteristics to predict treatment outcomes (13). The use of baseline characteristics only, currently limits these models to a moderate predictive accuracy. A major improvement might be to incorporate novel longitudinal risk parameters into dynamic models that can be updated during treatment and/or follow-up. Such dynamic models can serve personalized treatment choices, e.g. to distinguish in which patient a resection after CCRT needs to be considered. Optimization of radiotherapy by dose alteration With the theory that increasing the radiotherapy dose improves local control and OS, dose-escalation is an appealing option (14, 15). The excellent local control and OS reported for limited stage NSCLC patients treated with stereotactic ablative radiotherapy (SABR) (16) substantiates this theory. Safety and efficacy of dose escalation for LA-NSCLC was studied in several studies (9, 14, 17-20). A large phase III trial (RTOG-0617) (9) reported worse OS for the high dose arm (74 Gy, 2Gy fractions) compared to the standard arm (60 Gy, 2Gy fractions); 20.3 versus 28.7 months respectively. Furthermore, an increase of acute toxicity was seen in the dose- escalation arm. A recent Swedish randomized dose escalation phase II trial (19) (68 Gy versus maximum 84 Gy in 2 Gy per fraction) was prematurely terminated (N=36, 18 in each arm) due to excessive toxicity; 7 toxicity related death due to esophageal perforations and pneumonitis of which 5 in the dose-escalated arm and 2 in the standard arm. In both studies, dose escalation was performed by extending the overall treatment time. Since the outcomes of these recent trials, there is common