Session Item

Stereotactic re-irradiation (re-SRT) is an option for patients with recurrent glioblastoma (GBM). Unlike postoperative RT, where up to 2.5 cm margins may be added for potential microscopic spread, re-SRT typically uses 0 mm margins due risk of radionecrosis. The aims of this study were to investigate the radiological patterns of relapse after first and second irradiation. We also evaluated clinical outcomes and compared these with the literature.  

Selected patients with recurrent GBM received re-SRT with 10 x 3.5 Gy, 99% of the PTV received 100% dose and PTV=GTV. Volumes were contoured using a 1.5T MRI (Elements, Brainlab, Germany) and delivered with a Novalis STx (Varian, USA/ Brainlab). Radiological outcomes were assessed on successive Gd_T1 3D and FLAIR MRI sequences. Clinical endpoints were local control, overall survival and toxicity.

14 patients with recurrent GBM received re-SRT between 12/2015 and 08/2022. Eight patients were male, six were female and the median age was 61 yrs (41-77). All patients had previously received either resection followed by 30 x 2 Gy with concomitant temozolomide (TMZ) or 15 x 2.67 Gy with TMZ according to MGMT methylation status. 9/14 patients underwent re-resection and 7/14 had had at least one second-line systemic therapy prior to re-SRT. Of 10 evaluable patients, 8 recurrences were in-field having received a mean dose of 101.2% (97.1-103.3%) of the initial prescription, 1 was marginal (Dmean 90.1%, Dmin 56.9%, Dmax 104.3%) and 1 was distant (Dmean 33.9%, Dmin 24.7%, Dmax 59.7%). 5/10 had a second operation prior to re-SRT and anatomical shifts could not be excluded. The out of field recurrence did not show a 'leading edge' on FLAIR MRI prior to the new contrast-enhancing lesion 5 months after initial RT, nor did the marginal recurrence develop within the initial FLAIR signal. The median time to re-SRT following completion of adjuvant RT was 29 mths (7-224) and the median PTV at re-SRT was 6.6 cm3. Two patients received TMZ and two received bevacizumab concomitantly. Eleven patients had a follow-up MRI following re-SRT. Three patients had no evidence of tumour progression (mean follow-up 6 months, range 2.2-9.6). Five progessions were in-field, 1 was marginal and two were distant. Median progression-free survival following re-SRT was 5.8 mths (2.2-41.8). Five patients subsequently received bevacizumab (one combined with lomustine) and one was re-challenged with TMZ. The median overall survival was 8.3 months (1.5-92.5) and the six month overall survival rate was 79%, comparable to 73% reported in a meta-analysis of 50 papers. There were no documented related toxicities.

Median PFS, OS and toxicity in our series compared favourably with those reported in the literature. The pattern of relapse after re-SRT for GBM was similar to that described after first irradiation. Re-SRT with 10 x 3.5 Gy is a clinically meaningful option in selected patients with circumscribed recurrent GBM.