Spinal growth patterns after craniospinal irradiation in children with medulloblastoma.

PURPOSE: This study aimed to evaluate the impact on spine growth in children with medulloblastoma using either photon or electron craniospinal irradiation (CSI).

METHODS AND MATERIALS: This was a single institution retrospective review of children who were treated with CSI for medulloblastoma. Spine growth was measured on magnetic resonance imaging scans at defined locations on the basis of a published predictive model of spine growth after CSI. Differences between spine growth in the anterior, middle, and posterior aspect of the designated vertebral segments were also assessed. Differences between the groups treated with photons or electrons were assessed with student's t test.

RESULTS: A total of 19 patients (10 patients treated with electrons and 9 with photons) with a median follow-up time of 45.5 months (confidence interval, 34.9-55.1 months) were evaluated. Patients treated with electrons were younger than those who received photons (5.1 years [range, 3.8-9.0 years] vs 9.6 years [range, 3.5-12.9 years]); however, there were no differences in other clinical characteristics, treatment, or follow-up between the groups. Spine growth rate for patients treated with electrons fit the predictive model (104% ± 5.2%), but patients treated with photons had growth that was faster than predicted by the model (150% ± 47%) and different from that observed with electrons. The differences between treatment the modalities were statistically significant (P = .03). For patients treated with photons, there were no statistical differences between the growth rate of the anterior vertebral body compared with the posterior aspect, but for patients treated with electrons, a faster spine growth in the anterior L1-L5 lumbar spine was observed compared with the posterior lumbar spine (3.90 vs 2.52 mm/year; P = .006) without differences in the cervical or thoracic spine.

CONCLUSIONS: The use of electrons to treat the craniospinal axis in children with medulloblastoma resulted in no significant difference in spine growth compared with the predicted spine growth on the basis of previously published models using photons, but with a clinically insignificant faster spine growth rate in the anterior lumbar spine.