Translocator protein and methionine PET imaging in glioma

Introduction: Brain tumours continue to present specific imaging challenges, in particular related to grading and progression, from initial diagnosis, through surveillance and treatment planning, to response assessment and prognostication. In this project we explored the role of advanced MR biomarkers and PET imaging with two different tracers in glioma, [¹¹C]-Methionine and TSPO-binding ligand [¹¹C]-(R)-PK11195. Both PET tracers are ubiquitously low in the normal brain and high in the pathological tissue. Methods: In the first experimental chapter we investigated the role of [¹¹C]-(R)-PK11195 and relative cerebral blood volume (rCBV) in prediction of transformation in a cohort of 26 low grade gliomas (LGGs). Results were further validated on tissue derived from a consecutive cohort of 30 oligodendroglial tumours. The second step of our study extended to a larger cohort of 41 low and high grade gliomas, where we compared [¹¹C]-Methionine uptake to rCBV and diffusion tensor imaging (DTI) derived metrics in correlation to strongest signal areas of the tumour. Finally, in the last chapter we investigated the extent and potential causes of neuro-inflammation in the normal appearing brain of 66 patients with glioma, comparing to a cohort of 19 healthy controls and 27 patients with other types of intracranial tumours. Results: Progressively higher uptake of [¹¹C]-(R)PK11195 was seen in anaplastic gliomas compared to the lower grade (p=0.001), whilst rCBV failed to show a significant difference (p=0.23). Tissue data from TSPO antibody staining showed a very good correlation to the imaging results and were further validated in a cohort of 30 oligodendroglial tumours (sensitivity 92.85% and specificity 93.75%). In the second study, significant [¹¹C]-Methionine uptake was seen with increasing histological grade (p<0.001). Significant rank correlations were identified between peak methionine uptake ratios and peak values for both relative blood volume (p=0.004) and flow (p=0.036). Correlation of peak methionine values was additionally noted to inversely correlate with trough values of mean diffusivity (p=0.001). Tumours with vivid [¹¹C]-Methionine uptake appeared to have a poorer prognosis with shorter time-to-progression interval (p=0.015). Lastly, we found increased microglia activation, in both cerebral hemispheres in glioma patients compared to controls. This was most prominent in the tumour-bearing hemisphere (p<0.0001) but was also evident in the controlateral hemisphere (p=0.0078). Patients presenting with seizures showed a positive correlation between extra-tumoural microglial activation in the cerebrum and the duration of epilepsy. Conclusions: TSPO PET imaging can offer good accuracy in prediction of transformation which can be also translated in tissue analysis via antibody staining. Identifying tumour most malignant regions can be trivial using only MR derived data and PET imaging with [¹¹C]-Methionine can contribute to better surgical planning and understanding of tumour microenvironment. Increased neuro-inflammation in patients with glioma is linked to seizure activity, although it is difficult to ascertain cause and effect.