Abstract Book

S1221

ESTRO 37

do not assume to deliver the dose equally as planned. Instead the found difference identifies the need for novel QA methods that would allow the verification of adequate dose delivery to moving targets. These results form a starting point for discusssion. EP-2210 A review of QA submitted for the EORTC 1308 ROAM trial R. Brass 1 1 The Clatterbridge Cancer Centre, Physics, Bebington, United Kingdom Purpose or Objective To provide an overview of contouring and planning QA submissions for ROAM trial (radiotherapy vs observation following surgical resection of atypical meningioma: a randomised controlled trial). Material and Methods Two benchmark cases are available. One case is used to assess the PI's ability to contour targets and OARs according to trial guidelines. This case contains pre and post-op MR scans and a planning CT. The case is difficult as GTV abuts the orbital bone and, inferiorly, is divided by optic canal. ROAM guidelines dictate that the GTV-CTV margin is 5mm, extended to 10mm at dural edges. The other case is to assess the ability of the centre to plan according to the RT guidelines. The case is challenging as CTV approaches the patient surface, requiring the creation of a PTV_Report structure. PTV_Report consists of PTV cropped back from the surface by 5mm. The CTV overlaps with Optic apparatus, requiring a compromise of PTV coverage; the ipsilateral cochlea also sits within CTV. Benchmark QA for ROAM has been completed for 32 centres in 8 countries, with a further 3 centres partially submitted. Contouring benchmark reports are reviewed to investi- gate common deviations. Data from plan assessment forms is used to compare distributions for different treatment modalities. Results Of 35 contouring submissions, 32 were approved first time while 3 centres were asked to re-submit. Two of these subsequently passed while one is still to be re- submitted. Out of 37 total submissions, 10 were approved with no minor deviations. The remaining 24 submissions that passed were all returned with feedback due to one or more minor deviations. The most common area of feedback, as expected, was the delineation of the GTV and CTV. In particular, excluding the medial portion of GTV visible on the pre-op MR, extending the CTV margin at the dural edges and editing off uninvolved bone. Of the 33 planning cases approved, 7 failed the initial review. One further case has been rejected and is awaiting re-submission. Of the 8 rejections, 5 were wholly or partially due to the incorrect creation of PTV_Report. The others were all due to either missing OARs / PRVs or failing to keep the ipsilateral lens or cochlea within tolerance. The variety of treatment techniques permitted allowed for comparison of the distributions for each modality.

Conclusion The contouring standard for ROAM is high, with 2 rejections from 37 submissions; however the same deviations frequently occur. More guidance is required to ensure PIs understand what is required. A European contouring workshop has since been held to try to rectify this. Tomotherapy provides the most uniform coverage, with the highest D98% and lowest D1cc. Fixed field IMRT provides the greatest benefit to OARs, due to the increased degrees of freedom enabling the planner to avoid critical structures, however this at the expense of a uniform covering dose. NC VMAT offers more freedom than coplanar, allowing OARs to be avoided. Dose uniformity in the target is also maintained.