Abstract Book

S1067

ESTRO 37

contouring was based on interpolating between the cranial and caudal extent of an OAR, without considering the original clinical accurate contours (red contours in Figure). Individual clinical OAR contours in a RapidPlan TM model were combined to create composite salivary and swallowing structures. The simplified contours of each patient were assigned to the composite structures and the oral cavity of the model and used to create “simplified-contouring” knowledge-based plans (SC- plans). SC-plans were compared with their respective clinical plans, which were also automatically optimized, using accurate contours. Metrics for comparison were the boost planning target volume (PTV B )/elective PTV (PTV E ) receiving 95% and 107% of the prescription dose, mean dose to individual OARs using the accurately delineated clinical structures for dose reporting, and mean dose to the composite salivary/swallowing structures (comp sal /comp swal ).

Results In relation to OARs, when compared to the tangential plans, IMRT with SIB reduces the mean heart dose from 3,66 ±1,67 Gy to 1,99 ±1,04 Gy, and V 5 from 14.44 ±12.72 % to 5.96 ±4.10%. In addition, IMRT-SIB reduces also mean dose for left and right lung. On the other hand, IMRT-SIB increases the maximum dose to right breast, from 7.84 ±8.44 Gy to10.04 ±10.57Gy, and V 5 also increase from 0.15 ±0.34 % to0.40 ±0.87 %. Both techniques showed good target coverage, but IMRT better dose conformality.

Results Delineation of simplified-contours required an average of 8 minutes. On average, simplified contours included 95%/75%/93% of the individual salivary glands/oral cavity/swallowing structures. SC-plans were comparable to the clinical plans: composite salivary structures/oral cavity improved by an average of 1.1Gy/1.0Gy in the SC- plans, while dose to composite swallowing structures was an average 0.6Gy higher, largely attributable to higher doses in small-volume swallowing structures (Table). For instance upper esophageal sphincter/cricopharyngeal (smallest swallowing structures, on average) mean dose increased by 9.5Gy/7.5Gy for SC-plans over clinical plans.

Conclusion This SIB-IMRT technique helps in order to minimize doses to OARs in all the patients replanned, and decreases the total number of treatment fractions. Moreover the mean heart dose has been improved in all patients replanned. It is also important to take into account, that the use of IMRT techniques, with more Monitor Units and higher dose gradients, makes necessary to be much more careful with image guided systems. EP-1963 Is accurate contouring necessary for salivary and swallowing structure-sparing radiotherapy? A. Delaney 1 , M. Dahele 1 , B. Slotman 1 , W. Verbakel 1 1 VUMC, Cancer Center Amsterdam- Department of Radiation Oncology, Amsterdam, The Netherlands Purpose or Objective Current approaches to organ-at-risk (OAR) contouring emphasize accurate, anatomically faithful structures. Not only is this prone to inter-observer variation, but it is also time consuming and results in many individual structures which can make treatment plan optimization more challenging. This can create barriers to the implementation of advanced radiotherapy techniques, especially when resources are restricted. We therefore investigated if a simplified approach to salivary and swallowing structure contouring combined with automated plan optimization could lead to OAR sparing comparable to that of clinical locally-advanced head and neck cancer (HNC) plans, in which OARs were accurately and individually contoured. Material and Methods For 15 HNC patients, the oral cavity, parotid and submandibular glands and the swallowing structures were replaced by simple tube-shaped structures. Simplified