Leukoencephalopathy after prophylactic whole-brain irradiation with or without hippocampal sparing: a longitudinal magnetic resonance imaging analysis

Abstract

PURPOSE
Neurocognitive changes are well described after prophylactic or therapeutic whole-brain radiotherapy (WBRT) and have been reported as early as 3 months after radiotherapy (RT). Therefore, WBRT with protection of the hippocampal region (hippocampal avoidance, HA) has been proposed to preserve neurocognition. Our aim was to compare the risk of leukoencephalopathy after prophylactic cranial irradiation (PCI) with or without HA.
METHODS
Patients with small-cell lung cancer who received either lateral-opposed field PCI (non-HA-PCI; n = 9) or hippocampus avoidance PCI (HA-PCI; n = 9) with available magnetic resonance imaging (MRI) follow-up were identified and age matched. Pre-therapeutic and follow-up MRI after RT was analysed for leukoencephalopathy based on the Fazekas score. Bilateral cortical and subcortical brain structures were segmented and analysed for alterations in dosimetric parameters and volumes.
RESULTS
There was no significant difference of Fazekas scores between groups at baseline. Fazekas score differed in post-treatment with a median of 1 in the HA-PCI group and 2 in the non-HA-PCI group (p = 0.007). Significant increase of Fazekas score over time after RT was observed for HA-PCI patients (p = 0.001) but not for non-HA-PCI patients. Dmax (highest radiation dose) and brain volume receiving doses >25Gy were higher in HA-PCI patients. There were no significant volumetric differences for segmented brain structures between groups.
CONCLUSION
Radiological changes are more prominent after HA-PCI than after non-HA-PCI. Although no standardised neurocognitive testing was performed, the significantly increased Fazekas scores after HA-PCI are expected to interfere with neurocognitive function. Prospective long-term neurocognitive studies are warranted before HA-PCI is implemented in routine clinical practice.