Abstract Book

S1098

ESTRO 37

smaller effect on the SOBP than varying the HSG parameters separately. Still, an increasing dose was observed in the SOBPs in these cases. Hence, use of appropriate parameters in the MKM is important to obtain a flat plateau region in the SOBP. Parameter sensitivity analysis will further be performed using inhomogeneous patient anatomy. EP-2013 Predicting growth hormone deficiency after childhood cancer from hypothalamic-pituitary structures C. Stokkevåg 1 , L. Toussaint 2 , L. Muren 2 , C. Pedro 3 , R. Mikkelsen 4 , N. Birkebæk 5 , H. Schrøder 6 , Y. Lassen- Ramshad 6 1 Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway 2 Aarhus University Hospital, Department of Medical Physics, Aarhus, Denmark 3 Instituto Português de Oncologia de Lisboa Francisco Gentil, Department of Radiotherapy, Lisbon, Portugal 4 Aarhus University Hospital, Department of Neuroradiology, Aarhus, Denmark 5 Aarhus University Hospital, Paediatric Department, Aarhus, Denmark 6 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark Purpose or Objective Abnormalities in hypothalamic–pituitary axis (HPA) function are frequently seen in childhood cancer survivors after cranial radiotherapy (RT). Growth hormone (GH) deficiency (GHD) is usually the first and most frequent endocrine problem, with both incidence and severity increasing with time. There are known differences in tolerance doses to the hypothalamus and pituitary gland, but the combined effect and radio-sensitivity of small subunits of the HPA are less known. The aim of this study was to investigate feasibility of differentiating doses to HPA substructures and to explore predictive strategies of GHD based on radiation dose to individual and combined HPA substructures. Material and Methods Treatment plans of eleven paediatric patients (aged 4-11 years) receiving cranial photon RT with prescription doses of 40-60 Gy (1.6-1.8 Gy/fraction) were analysed. Only patients with tumors not invading the HPA were included. Five of the eleven patients were diagnosed with GHD after two year follow-up (patient VII-XI). Substructures of the HPA including the hypothalamus, pituitary gland, arcuate nucleus and the infundibulum were contoured on CT/MR images (Fig 1) and dose/volume data from the substructures and the full HPA were analysed. Level of GH and risk of GHD were estimated by three approaches: i) using a volume-based model (A) and a mean dose-based model (B) correlating peak growth hormone levels to hypothalamic dosimetry and time after RT (IJROBP 2002, JCO 2011); ii) estimating risk of GHD according to mean pituitary dose (model C) threshold at 30 Gy (J Pediatr 1991); and iii) adopting pragmatic radiosensitivity levels of the hypothalamus (16 Gy) or the pituitary gland (30 Gy) to both the arcuate nucleus and infundibulum to predict risk of GHD.

Results Mean doses for the substructures were in most cases comparable within a patient (Fig 2a), although the range exceeded 10 Gy for two patients and 5 Gy for four patients. Compared to the total HPA, the relative mean dose to the pituitary gland was 70-80% lower for two patients, with 10-20% higher relative mean dose for the hypothalamus. The hypothalamus model A predicted GHD in 10/11 patients at 1 year following RT, while model B predicted GHD in none of the patients at 1 year post RT, 4/11 at 2 y, and 8/11 at 3-5 y (Fig 2b). Including also pituitary model C resulted in risk profiles at 1y post RT at 55% and 2y-5y at 60-100%. The infundibular mean dose remained <30 Gy for three patients and <16 Gy for one patient, and correspondingly for two and one patient for the arcuate nucleus. The three approaches combined resulted in GHD in 10/11 patients at 5 y post RT, with the remaining patient at 70% risk of GHD.