Proton radiation therapy (PRT) for pediatric optic pathway gliomas: comparison with 3D planned conventional photons and a standard photon technique.

PURPOSE: Following adequate therapy, excellent long-term survival rates can be achieved for patients with optic pathway gliomas. Therefore, avoidance of treatment-related functional long-term sequelae is of utmost importance. Optimized sparing of normal tissue is of primary concern in the development of new treatment modalities. The present study compares proton radiation therapy (PRT) with a three-dimensional (3D)-planned multiport photon and a lateral beam photon technique for localized and extensive optic pathway tumors. METHODS AND MATERIALS: Between February 1992 and November 1997, seven children with optic pathway gliomas underwent PRT. For this study, we computed proton, 3D photon, and lateral photon plans based on the same CT data sets, and using the same treatment planning software for all plans. Radiation exposure for normal tissue and discrete organs at risk was quantified based on dose-volume histograms.
RESULTS: Gross tumor volume (GTV) ranged from 3.9 cm3 to 127.2 cm3. Conformity index (relation of encompassing isodose to GTV volume) was 2.3 for protons, 2.9 for 3D photons, and 7.3 for lateral photons. The relative increase of normal tissue (NT) encompassed at several isodose levels in relation to NT encompassed by the 95% proton isodose volume was computed. Relative NT volume of proton plan isodoses at the 95%, 90%, 80%, 50%, and 25% isodose level increased from 1 to 1.6, 2.8, 6.4, to a maximum of 13.3. Relative volumes for 3D photons were 1.6, 2.4, 3.8, 11.5, and 34.8. Lateral plan relative values were 6, 8.3, 11.5, 19.2, and 26.8. Analysis for small (<20 cm3) and larger (> 80 cm3) tumors showed that protons encompassed the smallest volumes of NT at all isodose levels. Comparable conformity and high-dose gradient were achieved for proton and 3D photon plans in small tumors. However, with increasing tumor volume and complexity, differences became larger. At the 50% isodose level, 3D photons were superior to lateral photons for small tumors; this advantage was equalized for larger tumors. At the lowest isodose level, 3D photons encompassed the highest amount of NT. Analysis of organs at risk showed that PRT reduced doses to the contralateral optic nerve by 47% and 77% compared to 3D photons and lateral photons, respectively. Reductions were also seen for the chiasm (11% and 16%) and pituitary gland (13% and 16%), with differences at clinically relevant tolerance levels. Furthermore, reduced dose exposure of both temporal lobes (sparing 39% and 54%) and frontal lobes was achieved with PRT.
CONCLUSION: PRT offered a high degree of conformity to target volumes and steep dose gradients, thus leading to substantial normal tissue sparing in high- and low-dose areas. It is expected that this will result in decreased long-term toxicity in the maturing child. Advantages of proton versus 3D photon plans became increasingly apparent with increasing target size and tumor complexity. Even in small tumors, conformity of 3D photon irradiation came at the expense of a larger amount of NT receiving moderate to low radiation doses. Lateral photons resulted in inferior dose distribution with high radiation exposure of clinically relevant normal tissues.