A new standard for coeliac plexus radiotherapy contouring.

699 Background: Patients with upper abdominal malignancies often experience severe, refractory abdominal pain caused by infiltration of the coeliac plexus. A pivotal trial reported objective alleviation of pain after delivery of stereotactic body radiotherapy (SBRT) 1 x 25 Gy to the coeliac plexus. However, a SBRT ‘surrogate’ volume was used as there is limited experience in delineating this structure. The aim of this study is to provide a computed tomography (CT) and magnetic resonance imaging (MRI) guided, peer reviewed, standard for coeliac plexus contouring to enhance radiotherapy accuracy and maximise patient benefit. Methods: Patients who received pancreatic radiotherapy between 2023 and 2025 at the Royal Marsden Hospital were eligible if they had a planning CT and upper abdominal MRI undertaken. Images from both modalities were fused. Systematic delineation of the coeliac plexus was done by three Radiation Oncologists and peer reviewed by a team of Radiologists. The CT/MRI guided coeliac plexus volumes were then compared with the SBRT surrogate volume (as defined in the 1 x 25 Gy trial) for each patient. Results: 60 patients had both an MRI and planning CT. 50 patients where the coeliac plexus was clearly visible were included. We identified that the plexus is located between the lower edge of T11 and mid-L2 vertebral bodies, with a median craniocaudal length of 24 mm (12-42) and a median volume of 2.4cc (0.6-5.7). The plexus is located between the diaphragm and adrenal gland on the left, and between the diaphragm and inferior vena cava or renal vessels on the right. A comparison of our CT/MRI guided volumes of the coeliac plexus with the SBRT surrogate volumes revealed a degree of geographical miss in all patients. The median volume of plexus covered by the SBRT surrogate clinical target volume (CTV) was 0.2cc (range 0-1.1cc), which corresponded to a median percentage volume of 90.8% (range 50-100%) that was not covered by the surrogate volume. The distributions of the geographical misses were as follows: ventral (10%), lateral (42%), ventral and lateral (48%). 10 patients were excluded from this study where the coeliac plexus was not visible due to little abdominal fat (n=3), no contrast enhancement used (n=3) or a large tumour on one side of the plexus (n=4), making it impossible to contour both sides. Conclusions: This is the first comprehensive, peer-reviewed, CT/MRI-guided study that identifies the location of the coeliac plexus for radiotherapy planning. We found that the plexus is visible in most patients. Where it is not clearly visible, using a 0.5cm margin around the aorta, as previously recommended, leads to geographical miss. Instead, we would recommend using a 0.5cm margin around the diaphragmatic crus, as this covers the area where the coeliac plexus is located. This data will inform artificial intelligence-based auto-contouring models of the coeliac plexus. Prospective evaluation and correlation with patient-reported outcomes is underway.