Dose-Constraint Model to Predict Neuroendocrine Dysfunction in Young Patients With Brain Tumors: Data From a Prospective Study.

Purpose: We report on a possible dose-constraint model to predict long-term neuroendocrine dysfunction after cranial irradiation in children and young adults with benign and low-grade brain tumors treated with stereotactic conformal radiation therapy (RT) in a prospective clinical trial.

Methods And Materials: Patients treated with stereotactic conformal RT (54 Gy in 30 fractions) were included for analysis if their co-registered planning computed tomography and magnetic resonance imaging scans were available, along with baseline and post-RT endocrine assessment for at least 2 years. The hypothalamus-pituitary axis (HPA) was contoured on the fused computed tomography-magnetic resonance imaging data set. Worsening of endocrine function was defined biochemically as a new onset endocrine deficit or worsening of preexisting endocrine deficit. Dosimetric indices of HPA, extracted using cumulative dose-volume histograms, were correlated with worsening endocrine function using logistic regression analysis.

Results: A total of 51 patients (median age: 13 years; range, 5-25 years) were included. Worsening post-RT endocrine levels were seen in 27 of 51 patients (47%). Growth hormone was the most commonly affected (70%), followed by cortisol (44%), gonadotropin (40%), and thyroxine (7%). The mean of the maximum and minimum doses to HPA was 42.1 Gy and 35.7 Gy, respectively. For patients with worsening endocrine levels, the mean maximum dose to HPA was 46.6 Gy compared with 36.5 Gy in patients with stable functions. The mean minimum dose to HPA was also higher (40.5 Gy vs 29.6 Gy) in patients with endocrine dysfunction. Logistic regression analysis identified the volume of HPA receiving 50% of the prescribed dose as the only statistically significant parameter predicting endocrine dysfunction. A dose of ≥27 Gy to any volume of HPA was associated with a 4-fold increase in risk of endocrine dysfunction (odds ratio: 4.05; 95% confidence interval, 1.07-15.62; P = .038).

Conclusions: Our prospective longitudinal study demonstrates the feasibility of HPA avoidance using modern, high-precision, conformal RT techniques and correlates HPA dosimetry with neuroendocrine dysfunction. We suggest restricting HPA doses to <27 Gy to minimize the risk of post-RT neuroendocrine deficits.