Pancreas: An Organ-at-Risk to Consider in Future Pediatric Abdominal Radiotherapy

Abstract

Irradiation to the pancreas during pediatric radiotherapy is associated with secondary late effects, yet the pancreas is not typically considered an organ-at-risk (OAR) during radiotherapy planning. This audit investigated incidental pancreatic doses to estimate the relative risk (RR) of developing diabetes in later life. Incidental pancreas doses were audited from 92 patients aged 2 to 19 historically treated with photons for craniospinal irradiation (CSI, N = 73) and for abdominal neuroblastoma (N = 19). Prescription doses ranged 21-36 Gy and 21-39 Gy (with boost to spine up to 50 Gy) for neuroblastoma and CSI patients, respectively. The pancreas was segmented on all planning CT scans following RTOG guidelines. Furthermore, the pancreas was split into its sub-volumes (head, body and tail) for abdominal cases only, since these scans were acquired with contrast enhancement. The RR of developing subsequent diabetes was estimated as a function of dose (D): RR = 1+0.65 × D × exp(-0.3 × D). Dose to the pancreas tail was used for neuroblastoma cases, whereas dose to the whole pancreas was used for CSI patients given that the pancreas could not be split in non-contrast scans. The mean dose delivered to the whole pancreas was 15.4±7.5 Gy for all patients, 12.7±4.7 Gy for CSI patients and 25.9±7.4 Gy for abdominal patients. Pancreas V10Gy was 61.2% for all patients, 51.5% for CSI, and 98.6% for abdominal patients; V30Gy = 0% in all cases. These doses represented an estimated diabetes RR of 7.3±1.5, 6.6±1.3, and 8.7±0.6, respectively; RR>5 in 88%, 85% and 100% of all patients, CSI and abdominal patients audited, respectively. Current incidental doses were associated with a diabetes RR>5 in all patients treated for abdominal neuroblastoma and the majority who received CSI. These findings suggest the pancreas should be routinely delineated and considered as an OAR in radiotherapy planning. This may help to identify higher risk patients and inform late-effect monitoring during survivorship. Further work in underway to assess if proton therapy could reduce the RR of diabetes particularly in abdominal cases. Improved availability of high-quality imaging during radiotherapy planning, such as contrast administration or planning MRI, is becoming more necessary to accurately delineate subsegments of the pancreas for improved RR estimations.