Delineation in thoracic oncology: a prospective study of the effect of training on contour variability and dosimetric consequences

INTRODUCTION As part of French residents' radiotherapy training, delineation workstations were available at a national teaching course. We report a prospective comparative study of a non small cell lung cancer (NSCLC) case delineated by 120 residents before and after a radioanatomy/radiotherapy lecture. MATERIALS AND METHODS The case of a patient with right upper lobe non small cell lung cancer (NSCLC) was provided for delineation to 32 groups of residents before and after a radiation therapy lecture about thoracic delineation. GTV, CTV and PTV was asked to each group. In a second step, the GTV, CTV and PTV were compared with those of 9 groups of senior physicians. Finally the consequences for treatment planning between each group before and after the course were explored. RESULTS The expert's average GTV, CTV and PTV were 89.1 cm3, 242.3 cm3 and 293.9 cm3 respectively. For residents, those volumes were 103.4 cm3, 242.3 cm3 and 457.9 cm3 before teaching, compared to 99.5 cm3, 224.2 cm3 and 412.5 cm3 after teaching. The overlap (OV) and kappa (KI) indices before and after education were respectively 0.58 and 0.73. Compared to senior physicians, OV and KI indices were lower in the residents group (p = 0.039 and p = 0.043). An increased dose to the lung is noted for the residents' dosimetry compared to the experts' (V20: 23.2% versus 36.5%) due to the larger PTV delineated. No significant difference was observed for other organs at risk. CONCLUSION There were no significant differences for the delineation of the GTV and CTV before and after the course, although the differences tended to decrease after the course. The good initial quality of the contours could explain the lack of difference. V20 for lung was higher in the residents group compared to the experts group (23.2% vs 36.5%). No other treatment planning consequences were observed for other critical organs.