12 Chapter 1 size next to smoking (a protective factor for OS) and the mean dose to the heart in their prediction model. This latter model has been validated in another lung cancer cohort as well as in an esophageal cancer cohort[17]. These externally validated prediction models are currently used for optimizing radiotherapy dose distributions and selecting patients for newer techniques like proton therapy in the Netherlands. Prediction models can be more reliable and reproducible, when you know more about the pathogenesis. This knowledge may help in the preselection of the DVH parameters of (subregions of) critical OARs while developing prediction models [15]. Besides, this may guide choosing the most relevant DVH-parameters in case of multicollinearity, as different anatomically closely related DVH parameters, like the heart, its subregions, and the lungs often have a strong correlation with each other. Summarizing, although prediction models suggest that irradiation of the heart leads to worse OS, a causal relationship between radiation dose to the heart and toxicity cannot be concluded from these papers. Based on the limited evidence available on radiation induced cardiac toxicity in esophageal cancer, we started this thesis to improve knowledge on the relevance and mechanisms of radiation induced cardiac toxicity as a first step towards NTCP modelling and to guide trade-offs between radiation dose to the heart and the lungs during the treatment planning process. Outline of this thesis Chapter 1 is the introduction of this thesis, in which provided background information on the literature at start of the thesis. Furthermore, we explain some of the terminology used and the end we highlight the unmet needs on cardiac toxicity in the treatment of esophageal cancer. In chapter 2 we performed a review of the available literature on radiation induced cardiac toxicity in the treatment of esophageal cancer. Chapter 3 consists of an editorial on radiation induced cardiac toxicity to increase awareness under other medical disciplines. Chapter 4 describes retrospective cohort study based on a population of 216 esophageal cancer patients who underwent curative radio(chemo)therapy to a relatively high radiation dose. We performed multivariable analyses combining clinical data with dose volume parameters to develop multivariable prediction models for heart as well as lung toxicity.
RkJQdWJsaXNoZXIy MTk4NDMw