Dose-Volume relationships and prognostic factors for brain radiation necrosis in high-dose brain Irradiation beyond 60 Gy

对于脑肿瘤以及伴有脑浸润或靠近颞叶的恶性肿瘤（如鼻咽癌），有效的局部控制受到正常组织毒性的阻碍。根据QUANTEC，超过60戈伊的毒性显著增加。Lawrence及其同事根据终点放射性坏死描述了正常组织毒性，而没有NTCP建模。他们的尝试非常简单，只考虑了各自DVH的最大剂量，将其转换为BED值，并绘制BED与辐射坏死率的关系图。根据一般的NTCP模型，基于未知EUD参数对曲线进行建模，该参数将通过最大似然方法进行拟合-这将能够导出高-剂量体积对放射性坏死发展的影响（如，e. G.放射外科可以向大脑输送更高的剂量，而不会造成明显的损伤，这可以作为体积效应的进一步论据）。实际上，可能有更多的参数影响脑坏死的发展（将根据症状进行分层），例如照射体积的大小、伴随的全身治疗、肿瘤的定位和许多其他因素;除了剂量-体积相关性外，这些因素的影响将通过多变量logistic回归分析进行评估。在本研究项目中，连续计划回顾性识别此类患者，评估其质子计划并开发脑放射性坏死的NTCP模型。

Effective local control in brain tumors as well as in malignant neoplasias with brain infiltration or in close vicinity to the temporal lobe such as in nasopharyngeal cancer is hampered by normal tissue toxicity.There is a considerable increase in toxicity beyond 60 Gy according to QUANTEC.Lawrence and colleagues have described normal tissue toxicity according to the endpoint radiation necrosis without NTCP modelling. Their attempt was quite straightforward only considering the maximum dose of their respective DVH, transferring this into a BED value and plotting BED versus the rate of radiation necrosis.This method contains the disadvantage not considering the whole texture of the DVH which can be done by employing EUDs; according to the general NTCP model, a curve is modelled based on unknown EUD parameters which will be fitted by maximum likelihood methods - this will enable to derive the volumetric influence of high-dose volumes on the development of radiation necrosis (as, e. g. radiosurgery can deliver far higher doses to the brain without causing significant damage which can be taken as further argument for a volume effect).In reality, there might be much more parameters influencing the development of brain necrosis (stratification will be done by symptoms) such as the size of the irradiated volume, concomitant systemic therapy, localization of the tumor and many more factors; the influence of these factors besides the dose-volumetric association will be assessed by multivariate logistic regression analysis.Within this research project, it is consecutively planned to retrospectively identify such patients, to evaluate their proton plans and to develop a NTCP model on brain radiation necrosis.