P16.29 POST-OPERATIVE DIFFUSION TENSOR IMAGING IDENTIFIES RESIDUAL INVASIVE TUMOR FOLLOWING MAXIMAL RESECTION OF GLIOBLASTOMAS AND PREDICTS LOCATION OF TUMOUR PROGRESSION

INTRODUCTION: Although techniques such as 5-aminolevulinic acid (5-ALA) guided resections can increase extent of 'maximal' resections the invasive nature of glioblastomas makes total resections impossible and residual disease inevitable. Previous work decomposing diffusion tensor imaging (DTI) data into isotropic (p) and anisotropic (q) components identifies an abnormality that has been shown to be invasive tumor on biopsy. This study aims to identify the invasive residual tumor left behind after attempted maximal resection and see if this can predict areas of tumor progression. METHODS: 15 patients (mean age 55.2; 11 males) with glioblastomas were imaged both pre-operatively and within 72 hours of a 5-ALA guided resection. Imaging involved standard anatomical sequences (to assess the extent of resection) as well as a DTI. Maps of p and q were coregistered to the post-contrast T1-weighted axial images and abnormal areas of the q (gross tumor) and p (tumor and invasive margin) were delineated and the difference was taken as the invasive residual tumor. All patients were treated with radiotherapy and concomitant and adjuvant temozolomide chemotherapy and followed up until progression or a minimum of one year. Contrast enhanced T1-weighted images of progressive disease were used to identify the pattern of recurrence. This recurrent disease was compared with the outlined invasive residual tumor from the post-operative DTI. RESULTS: Post-operative imaging within 72 hours of surgery identified a complete resection of the contrast-enhancing tumor (CRET) in 9 patients (60%). At the time of analysis 6 had progressed with a median time to progression of 444 days. The pattern of recurrence was within the DTI-defined invasive residual area in 5 of the 6 patients progressing (sensitivity 83%). In 3 cases the enhancing progressive tumor was >2cm from the edge of the resection cavity. Of the 6 patients undergoing partial resection of the enhancing tumor (PRET) 5 had progressed with a median time to progression of 258 days. In all cases progression occurred in the area of residual contrast enhancing disease. CONCLUSION: Imaging the residual invasive tumor using DTI can identify sites of tumor progression. These regions might be a target for further treatment options including extended resections, directing convection-enhanced delivery or radiotherapy planning. Progression at sites of residual contrast enhancement suggests we need to consider second-look surgery in selecte