MON-458 Radiological Alterations Of The Hypothalamic-pituitary Region After Craniospinal Irradiation For Medulloblastoma During Childhood.

Background Hypothalamic-pituitary (HP) dysfunction after craniospinal irradiation (CSI) is frequently observed in childhood brain tumor survivors (CBTS). The risk increases with higher dose and younger age at time of radiation exposure. Whether radiological alterations of the HP-region or alterations in brain volume are predictive of HP dysfunction remains unknown. Aim The aim of this (retrospective) study was to quantify radiological alterations of the HP-region in time in CBTS after CSI and to correlate these alterations to the occurrence of HP dysfunction. Methods Ninety survivors of childhood medulloblastoma (mean age at diagnosis: 8.3 years) from a previously reported nationwide cohort, treated with CSI between 2002-2012, were included. Sixty CBTS (67%) were diagnosed with HP dysfunction during follow-up (FU). All MRI scans were collected from time of diagnosis, post-neurosurgical intervention, post-radiation and 3 and 5 years of FU. Pituitary height (PH) and width (PW) were measured on mid-sagittal images. The pituitary stalk (PS) width was assessed by measuring the ratio of the PS to basilar artery (BA) on axial images on the same plane in the middle of the PS. Volume measurements of the PG (PGV) were performed. Observers were blinded for outcome of HP function. Separate analyses were performed for children < 6 years at time of CSI. All measurements were corrected for age and gender using Z-scores. Statistical analyses were performed by SPSS (General Linear Model repeated measures with Bonferroni correction). Results Mean PH (p<0.01, 95% CI [0.25, 1.27], PW (p<0.01, 95% CI [0.85-2.32], width of the PS (p<0.01, 95% CI [0.38-3.0] and PGV (p<0.05, 95% CI [0.12-1.29] declined during FU. No correlation was observed between radiological alterations in time and the presence of HP dysfunction. On a marginal trend level, an interaction effect was seen for PS/BA ratio and the presence of HP dysfunction (p= 0.097, F= 2.44). HP dysfunction occurred more frequently in CBTS treated with CSI < 6 years (71% vs 58%). Radiological alterations in this age group could also not be related to occurrence of HP dysfunction. Conclusion When compared to age and gender reference values, CBTS treated with CSI show radiological alterations of the HP region in time. Exposure to cranial irradiation in childhood seems to have a negative effect on PH, PW, width of PS and PGV. However, these effects cannot be related to the development of HP dysfunction or age at time of CRT. Prospective studies are needed to confirm these results, with systematic evaluation of the HP function in time.