Abstract
BACKGROUND: Whole brain radiotherapy (WBRT) has been the radiation treatment of choice for patients with primary central nervous system lymphoma (PCNSL). Neurocognitive functions (NCFs) decline is surely one of the main significant complications which may occur. It has been widely shown that neurocognitive decline is mainly caused by the radiation-induced injury to the hippocampus. The conformal hippocampal avoidance during the course of WBRT might provide meaningful NCFs preservation. Our study aims to investigate the correlation between the site of (Primary Central Nervous System Lymphoma) PCNSL lesions and the hippocampal region in order to explore the feasibility of routinely sparing of the hippocampus during a WBRT to prevent neurocognitive decline. MATERIAL AND METHODS: Patients (> 18 years, ECOG <4) with pathologically proven PCNSL and MRI image pre-treatment were retrospectively reviewed. All patients had received high-dose of methotrexate (HD-MTX) before WBRT. T1-weighted, post-contrast axial MR image sets obtained prior to cranial irradiation were imported on Varian Eclipse treatment planning system, version 11 (Varian Medical Systems) and registered with the simulation CT. The hippocampus as well as each PCNSL lesions were contoured. Three dimensional envelopes surrounding the hippocampus were generated adding 5, 10, and 15 mm and the distance of brain lesions were recorded as <5 mm, 5 to <10 mm, 10 to <15 mm, and >15 mm from the hippocampus. The minimum margin of 5 mm was taken into account for systematic setup error and dose fall-off between whole brain clinical target volume and the hippocampus. RESULTS: Thirty-eight patients were treated between 2005 and 2018 and 36 pts with 57 lesions were eligible for this study. PCNSL lesions’ locations were: deep brain structures (26%), parietal lobe (23%), frontal lobe (19%), temporal lobe (14%), occipital lobe (7%), brainstem (5%), other sites (6%). Nine patients were affected by multiple lesions. In 18/57 lesions (31.6%) the distance from the hippocampus region was less than 5 mm and seven of them (12.3%) involved the hippocampus. Lesions over 15 mm from the hippocampus were observed in 30 cases (52.6%), while only the 15,8% was between 5 and 15 mm. CONCLUSION: These data show that routinely sparing of the hippocampus region is not recommended. Anyway, it should be considered in selected patients, when the spatial distribution of PCNSL is far more than 5 mm from the lesion.