Introduction Surgical failure of local control is a feature of GBMs. Resecting non-enhancing tumour improves survival. PRaM-GBM, a multicentre imaging biomarker study, has validated diffusion tissue signatures to predict sites of GBM progression. Using this data, we have undertaken an IDEAL Stage 0 study (to translate this biomarker into the OR) and a Stage 1 study to provide proof of concept of targeting non-enhancing tumours.

Methods 8 patients underwent DTI imaging (predicting sites of progression), [C-11]-methionine PET (surrogate of tumour resected with 5-ALA fluorescence-guidance) and contrast-enhanced tumour. The q area outside increased PET uptake was taken as the surgical target. These regions were incorporated into image guidance, and biopsies were taken to confirm tumour presence (Stage 0 Study). Finally, the Stage 1 study – in a patient resection was extended into the q region.

Results Stage 0 study developed the method of identifying residual tumour in patients intra-operatively. Biopsies confirmed these regions contained high-density tumours. Extended resection was deemed safe in 5 patients. A single patient with a non-dominant frontal lobe tumour safely underwent extended resection 2cm into the q region posterior to the enhancing tumour.

Conclusions We have defined an imaging biomarker for non-enhancing tumours. Biopsies have confirmed it is tumour and demonstrated we can extend resection. Further studies to demonstrate efficacy are required.