A prospective comparison of adaptive and fixed boost plans in radiotherapy for glioblastoma

PURPOSE: To analyze the efficacy of adaptive radiotherapy (ART) for glioblastoma. METHODS: Sixty-one glioblastoma patients who received ART were prospectively evaluated. The initial clinical target volume (CTVinitial) was represented by T2 hyperintensity on postoperative MRIs (pre-RT MRI [MRIpre])plus 10 mm. The initial planning target volume (PTVinitial) was the CTVinitial plus a 5-mm margin. The PTVinitial received 40 Gy. An MRI and a second planning CT were performed during radiotherapy (MRImid). Two types of boost CTVs (the resection cavity and residual tumor on enhanced T1-weighted MRI plus 10 mm) were created based on the MRIpre and MRImid (CTVboost-pre and -mid). The boost PTV (PTVboost) was the CTVboost plus 5 mm. Two types of boost plans (fixed and adaptive boost plans in the first and second planning CT, respectively) of 20 Gy were created. The PTV based on the post-RT MRI (PTVboost-post) was created, and the dose-volume histograms of the PTVboost-post in the fixed and adaptive boost plans were compared. Additionally, the conformity indices (CIs) of the fixed and adaptive boost plans were compared. RESULTS: The median V95 of the PTVboost-post of the fixed and adaptive boost plans (V95pre and V95mid) were 95.6% and 98.3%, respectively (P < 0.01). The median V95pre and V95mid of patients after gross total resection (GTR) were 97.4% and 98.8%, respectively (P = 0.41); in contrast, the median values of patients after non-GTR were 91.9% and 98.2%, respectively (P < 0.01). The median CIs of the fixed and adaptive boost plans in all patients were 1.45 and 1.47, respectively (P = 0.31). The median CIs of the fixed and adaptive boost plans in patients after GTR were 1.61 and 1.48, respectively (P = 0.01); in contrast, those in patients after non-GTR were 1.36 and 1.44, respectively (P = 0.13). CONCLUSION: ART for glioblastoma improved the target coverage and dose reduction for the normal brain. By analyzing the results according to the resection rate, we can expect a decrease in normal brain dose in patients with GTR and an increase in coverage in those with partial resection or biopsy.