To analyze the recurrence patterns in patients with newly diagnosed glioblastoma (GBM) treated with conformal radiotherapy (RT) plus concomitant and adjuvant temozolomide (TMZ), and to compare the patterns of failure according to different target volume delineations.

One hundred and five patients with GBM which recurred after three-dimensional (3D) conformal RT plus TMZ were evaluated. The clinical target volume (CTV) used for our treatment planning (S'Andrea plans) consisted of residual tumor and resection cavity plus 2-cm margins according to recent randomized trials of the European Organisation for Research and Treatment of Cancer (EORTC). MRI scans showing tumor recurrences were fused with the planning computed tomography (CT), and the patterns of failure were analyzed dosimetrically using dose-volume histograms. For each patient a theoretical plan based on the addition of postoperative edema plus 2-cm margins according to current guidelines of Radiation Therapy Oncology Group (RTOG) was created and patterns of failure were evaluated.

The median overall survival and progression-free survival were 14.2 months and 7.5 months, respectively. Recurrences were central in 79 patients, in-field in 6 patients, marginal in 6 patients, and distant in 14 patients. Analysis of O(6)-methylguanine-DNA-methyltransferase (MGMT) promoter methylation status showed different recurrence patterns of GBMs in patients with MGMT methylated compared with patients with MGMT unmethylated status. Recurrences occurred central/in-field and outside in 64% and 31% of methylated patients, and in 91% and 5.4% of unmethylated patients, respectively (P=0.01). Patterns of failure were similar between the different treatment plans, however the median volume percent of brain irradiated to high doses was significantly smaller for our plans than for RTOG plans (P=0.0001).

Most of patients treated with RT plus concomitant and adjuvant RT have central recurrences, however distant new lesions may occur in more than 10% of patients. The use of target delineation using postoperative residual tumor and cavity plus 2-cm margins is associated with smaller volumes of normal brain irradiated to high doses as compared with plans including expanded edema, without a significant increase of the risk of marginal recurrences. Future clinical randomized studies need to compare the different planning methods in terms of efficacy and risk of late radiation-induced toxicity.