Protons in posterior fossa ependymoma- a dosimetric comparison with photons

Abstract

Purpose or Objective Adjuvant radiotherapy in ependymoma improves both local control and survival outcomes. The 5 year overall survival in paediatric ependymoma is around 85%. With such outcomes, the late effects of treatment assume prime importance. Protons when compared with photons have the potential to spare organs at risk (OARs), and thereby decrease late complications. This study was designed to compare the dosimetry of proton with photon radiotherapy plans in patients with posterior fossa ependymoma, in order to understand the potential for reduced late effects. Materials and Methods Data were extracted for ten patients with posterior fossa ependymoma treated with protons who had Volumetric Modulated Arc Therapy (VMAT) photon plans prepared in case of cyclotron downtime. Five were single phase plans to a total dose of 59.4Gy in 33 fractions, with appropriate limitation of dose at spinal cord level, and 5 were two phase plans, to 54Gy in 30 fractions followed by a boost of 5.4Gy in 3 fractions to the volume above the foramen magnum. Clinically relevant dosimetric variables of target and routinely outlined OARs for data collection were selected based on our clinical practice and with reference to the European Particle Therapy Network (EPTN) consensus on radiation dose constraints for OARs in neuro-oncology. Mean values were compared between proton and photon plans using a two tailed paired sample t-test. P values of less than 0.05 were considered statistically significant. Results Of ten patients, 7 were females. The median age was 7 years (range 3 to 17 years). For almost all dosimetric variables, proton plans showed a statistically significant dosimetric advantage in OAR variables, including to the optic nerves and chiasm, cochleae, pituitary gland, hypothalamus, hippocampi, temporal lobes, brain and brainstem. The mean D0.03cc for brainstem, skin and spinal cord were not statistically significantly different between modalities (Figure 1). The mean integral dose was significantly lower in protons compared to photons (p=0.00) with a median proton to photon integral dose ratio of 0.55 (range-0.41 to 0.74). The mean percentage of the PTV receiving 95% of prescription dose and the mean homogeneity index were not statistically significantly different between proton and photon plans (p=0.11 and p=0.98 respectively. Conclusion This study showed marked dosimetric advantage for protons in OAR sparing and lowered integral doses, while maintaining similar target volume coverage. This is anticipated to reduce the risk of late effects, particularly hearing loss, endocrine dysfunction, neurocognitive effects, and second malignancies, while maintaining the established high rates of local control with photons.