Radiation dosimetry predicts IQ after conformal radiation therapy in pediatric patients with localized ependymoma

Abstract

To assess the effects of radiation dose-volume distribution on the trajectory of IQ development after conformal radiation therapy (CRT) in pediatric patients with ependymoma. The study included 88 patients (median age, 2.8 years +/- 4.5 years) with localized ependymoma who received CRT (54-59.4 Gy) that used a 1-cm margin on the postoperative tumor bed. Patients were evaluated with tests that included IQ measures at baseline (before CRT) and at 6, 12, 24, 36, 48, and 60 months. Differential dose-volume histograms (DVH) were derived for total-brain, supratentorial-brain, and right and left temporal-lobe volumes. The data were partitioned into three dose intervals and integrated to create variables that represent the fractional volume that received dose over the specified intervals (e.g., V(0-20 Gy), V(20-40 Gy), V(40-65 Gy)) and modeled with clinical variables to develop a regression equation to estimate IQ after CRT. A total of 327 IQ tests were performed in 66 patients with infratentorial tumors and 20 with supratentorial tumors. The median follow-up was 29.4 months. For all patients, IQ was best estimated by age (years) at CRT; percent volume of the supratentorial brain that received doses between 0 and 20 Gy, 20 and 40 Gy, and 40 and 65 Gy; and time (months) after CRT. Age contributed significantly to the intercept (p > 0.0001), and the dose-volume coefficients were statistically significant (V(0-20 Gy), p = 0.01; V(20-40 Gy), p < 0.001; V(40-65 Gy), p = 0.04). A similar model was developed exclusively for patients with infratentorial tumors but not supratentorial tumors. Radiation dosimetry can be used to predict IQ after CRT in patients with localized ependymoma. The specificity of models may be enhanced by grouping according to tumor location.