Margriet Kwint

General discussion and future perspectives 163 8 currently ongoing RTOG 1308 phase III trial has as primary objective to compare OS, lymphopenia and cardiac toxicity of LA-NSCLC patients, who are randomized between IMRT and proton beam therapy (93). Secondary endpoints of this trial are 2-year PFS, grade ≥3 adverse events, quality of life and cost effectiveness. In this trial IMPT and isotoxic dose escalation (60 to 70 Gy) are allowed. This is important, because the potentially improved local control of dosimetric advantages in the proton-arm due to dose escalation by OAR sparing can be assessed. Hopefully this trial will contribute to our knowledge on the clinical benefit of IMPT compared to the widely used IMRT. The use of NTCP-models supports personalized radiotherapy. For example in patients with a low toxicity risk, dose escalation can be applied, and patient with a high toxicity risk can be selected for more conformal radiation techniques such as IMPT. Because of the limited availability of proton therapy in the Netherlands and high costs, it is important to select those patients that will benefit the most from this treatment. Therefore, in the Netherlands a so called “model based approach” is implemented, to decide which lung cancer patients should be selected for proton therapy (91). In this approach, NTCP models of radiation pneumonitis, acute esophagitis and mortality are used to estimate what the expected difference (ΔNTCP) in complication risk is between photons and protons. If this comparison reveals a considerable decrease of this complication risk, proton therapy is indicated. These NTCP-models are based on photon radiotherapy data, hence validation with proton radiotherapy data is essential in the future. In summary, new radiotherapy techniques such as IMPT might lead to an improved sparing of OAR and are therefore potentially suitable for dose escalation with improved outcome for the patient. Conclusions In the last decades big steps have been made in the treatment for LA-NSCLC. The studies presented in this thesis all contribute to the optimization of radiotherapy for LA-NSCLC patients to reduce toxicity while improving locoregional control. Several improvements in the personalized treatment of lung cancer patients are expected to be implemented within clinical practice in the coming years. Cooperation between different research areas such as pharmacology, biostatistics, artificial intelligence, radiology, medical physics and radiation oncology, is essential to achieve these improvements.