Normal Tissue Complication Probability Model for Radiation-induced Temporal Lobe Injury after Intensity-modulated Radiation Therapy for Nasopharyngeal Carcinoma

Abstract

PURPOSE: To identify predictors for the development of temporal lobe injury (TLI) after intensity-modulated radiation therapy (IMRT) for nasopharyngeal carcinoma. MATERIALS AND METHODS: Data in 351 patients with nasopharyngeal carcinoma treated with IMRT were reviewed retrospectively according to institutional ethics committee approval. Clinical factors associated with TLI were analyzed. Dose-volume histograms for 550 evaluable temporal lobes were analyzed, and the predictive value of therapy-associated and patient-associated factors for the occurrence of TLI was evaluated. Survival curves were depicted by using the Kaplan-Meier method and compared by using the log-rank test. Logistic regression analysis was used for multivariate analyses. RESULTS: Median follow-up was 76 months (range, 6-100 months). Twenty-nine of 351 patients (8.3%) developed TLI; 21 patients had unilateral TLI, and eight had bilateral TLI. Median latency from IMRT until first TLI was 33 months (range, 12-83 months) among patients with TLI. The actuarial TLI-free survival rates were 94.4% and 91.3% at 3 and 5 years after radiation therapy, respectively. Logistic regression analysis demonstrated that dose delivered to a 1-cm(3) volume of the temporal lobe (D1cc) was the only independent predictor for TLI. The biologically equivalent tolerance doses at 2 Gy for a 5% and 50% probability of developing TLI were 62.83-Gy equivalents (95% confidence interval: 59.68, 65.97) and 77.58-Gy equivalents (95% confidence interval: 74.85, 80.32), respectively. CONCLUSION: D1cc is predictive for radiation-induced TLI, suggesting that delivery of a high dose of radiation to a small volume of the temporal lobe is unsafe. A D1cc of 62.83 Gy by using a correction formula for varying fraction size may be the dose tolerance of the temporal lobe.