Radiation Oncology (Sep 2019)

MRI based neuroanatomical segmentation in breast cancer patients: leptomeningeal carcinomatosis vs. oligometastatic brain disease vs. multimetastastic brain disease

  • Michael Mayinger,
  • Antonia Reibelt,
  • Kai Joachim Borm,
  • Johannes Ettl,
  • Jan J. Wilkens,
  • Stephanie Elisabeth Combs,
  • Markus Oechsner,
  • Marciana Nona Duma

DOI
https://doi.org/10.1186/s13014-019-1380-3
Journal volume & issue
Vol. 14, no. 1
pp. 1 – 7

Abstract

Read online

Abstract Purpose Pathogenesis of brain metastases/meningeal cancer and the emotional and neurological outcomes are not yet well understood. The hypothesis of our study is that patients with leptomeningeal cancer show volumetric differences in brain substructures compared to patients with cerebral metastases. Methods Three groups consisting of female breast cancer patients prior to brain radiotherapy were compared. Leptomeningeal cancer patients (LMC Group), oligometastatic patients (1–3 brain metastases) prior to radiosurgery (OMRS Group) and patients prior to whole brain radiation (WB Group) were included. All patients had MRI imaging before treatment. T1 MRI sequences were segmented using automatic segmentation. For each patient, 14 bilateral and 11 central/median subcortical structures were tested. Overall 1127 structures were analyzed and compared between groups using age matched two-sided t-tests. Results The average age of patients in the OMRS group was 60.8 years (± 14.7), 65.3 (± 10.3) in the LMC group and 62.6 (± 10.2) in the WB group. LMC patients showed a significantly larger fourth ventricle compared to OMRS (p = 0.001) and WB (p = 0.003). The central corpus callosum appeared smaller in the LMC group (LMC vs OMRS p = 0.01; LMC vs WB p = 0.026). The right amygdala in the WB group appeared larger compared with the OMRS (p = 0.035). Conclusions Differences in the size of brain substructures of the three groups were found. The results appear promising and should be taken into account for further prospective studies also involving healthy controls. The volumetrically determined size of the fourth ventricle might be a helpful diagnostic marker in the future.