11 General introduction Radiation induced cardiac toxicity. Most clinical publications on radiation-induced cardiac toxicity were published after 2015. Historically, cardiac toxicity was considered a (very) late event and increased rates of cardiovascular diseases were observed in long term survivors of breast cancer and Hodgkin lymphoma[8,9]. Incidence rates of myocardial infarction (HR: 1.22(95%CI:1.06-1.42)), pericarditis (HR: 1.61(95%CI: 1.06-2.43)) as well as valve disorders (HR: 1.54(95%CI: 1.11-2.13)) were higher in left-sided breast cancer patients that were treated with radiotherapy as compared to right-sided breast cancer patients. For Hodgkin lymphoma survivors, an increased rate of heart failure (4.9 times) as well as myocardial infarctions (3.6 times) has been observed as compared to normal populations. The combination of radiotherapy and chemotherapy (anthracyclines) resulted in the highest risk of late cardiac toxicity in this population[10]. Literature on cardiac toxicity and radiation dose parameters is limited. The only externally validated NTCP model originates from breast cancer patients. In 2013, Darby et al published this data. The rate of major coronary events increased linearly with the mean radiation dose to the heart by 7.4 % per Gray[11]. These data were validated in another breast cancer population by Boogaard et al [12]. Overall survival as a surrogate for toxicity endpoints? As causes of death are often unknown, overall survival (OS) and/or death have been used as alternative endpoints in relation to normal tissue dose. This probably originated from trials were better tumor specific survival or local control was observed, while OS rates did not improve or even worsened[13,14]. Analyzing OS as an endpoint instead of toxicity is attractive because it can combine toxicities of, for example lung and heart, while considering tumor prognostic factors, like tumor stage. Moreover, it is a truly relevant endpoint. However, OS does not provide information on causal relationships. In addition, OS is less sensitive and provides less guidance on which specific regions of the thoracic region you should try to spare in radiotherapy treatment planning. In the treatment of intrathoracic tumors, several prediction models for OS have been developed [5,15,16]. These models include tumor-specific prognostic factors like tumor size or lymph node status next to DVH-parameters of the heart. Only Speirs et included a lung DVH parameter next to a heart DVH parameter in their final multivariable prediction model. The model of Defraene et al. included tumor 1
RkJQdWJsaXNoZXIy MTk4NDMw