Estimation of the probability of mediastinal involvement: a statistical definition of the clinical target volume for 3-dimensional conformal radiotherapy in non-small-cell lung cancer?

Abstract

Conformal irradiation of non-small cell lung carcinoma (NSCLC) is largely based on a precise definition of the nodal clinical target volume (CTVn). The reduction of the number of nodal stations to be irradiated would render tumor dose escalation more achievable. The aim of this work was to design an mathematical tool based on documented data, that would predict the risk of metastatic involvement for each nodal station. From the large surgical series published in the literature we looked at the main pre-treatment parameters that modify the risk of nodal invasion. The probability of involvement for the 17 nodal stations described by the American Thoracic Society (ATS) was computed from all these publications and then weighted according to the French epidemiological data. Starting from the primitive location of the tumour as the main characteristic, we built a probabilistic tree for each nodal station representing the risk distribution as a function of each tumor feature. From the statistical point of view, we used the inversion of probability trees method described by Weinstein and Feinberg. Taking into account all the different parameters of the pre-treatment staging relative to each level of the ATS map brings up to 20,000 different combinations. The first chosen parameters in the tree were, depending on the tumour location, the histological classification, the metastatic stage, the nodal stage weighted in function of the sensitivity and specificity of the diagnostic examination used (PET scan, CAT scan) and the tumoral stage. A software is proposed to compute a predicted probability of involvement of each nodal station for any given clinical presentation. To better define the CTVn in NSCLC 3DRT, we propose a software that evaluates the mediastinal nodal involvement risk from easily accessible individual pre-treatment parameters.