Pattern of relapse of glioblastoma multiforme treated with radical radio-chemotherapy: Could a margin reduction be proposed?

To analyse the pattern of recurrence of patients treated with Stupp protocol in relation to technique, to compare in silico plans with reduced margin (1 cm) with the original ones and to analyse toxicity. 105 patients were treated: 85 had local recurrence and 68 of them were analysed. Recurrence was considered in field, marginal and distant if >80 %, 20-80 % or <20 % of the relapse volume was included in the 95 %-isodose. In silico plans were retrospectively recalculated using the same technique, fields angles and treatment planning system of the original ones. The pattern of recurrence was in field, marginal and distant in 88, 10 and 2 % respectively and was similar in in silico plans. The margin reduction appears to spare 100 cc of healthy brain by 57 Gy-volume (p = 0.02). The target coverage was worse in standard plans (pt student < 0.001), especially if the tumour was near to organs at risk (pχ2 < 0.001). PTV coverage was better with IMRT and helical-IMRT, than conformal-3D (pAnova test = 0.038). This difference was no more significant with in silico planning. A higher incidence of asthenia and leuko-encephalopathy was observed in patients with greater percentage of healthy brain included in 57 Gy-volume. No differences in the pattern of recurrence according to margins were found. The margin reduction determines sparing of healthy brain and could possibly reduce the incidence of late toxicity. Margin reduction could allow to use less sophisticated techniques, ensuring appropriate target coverage, and the choice of more costly techniques could be reserved to selected cases.