Cranial neuronavigation with direct integration of (11)C methionine positron emission tomography (PET) data -- results of a pilot study in 32 surgical cases.

Abstract

MRI detects small intracranial lesions, but has difficulties in differentiating between tumour, gliosis and edema. (11)C methionine-PET may help to overcome this problem. For its appropriate intra-operative use, it must be integrated into neuronavigation. We present the results of our pilot study with this method. 32 patients with 34 intracranial lesions detected by MRI underwent additional (11)C methionine-PET, because the pathophysiological behaviour or the tumour delineation was unclear. All lesions were treated surgically. In 25 patients PET data could be integrated directly into cranial neuronavigation. (11)C methionine uptake was observed in 27/34 lesions, 26 of them were tumours: 14 malignant and 7 benign gliomas, 3 gliomas without further histological typing, one Ewing sarcoma and one non-Hodgkin lymphoma. Only one (11)C methionine positive lesion was non-tumourous: it was staged as post-irradiation necrosis in a patient operated on for a malignant glioma. 3/7 (11)C-methionine negative lesions were classified as gliosis (n=2) and M. Whipple (n=1), but 4/7 were tumours: 2 astrocytomas WHO(degrees)II, 1 DNT and one astrocytoma WHO(degrees)III. The sensitivity of (11)C methionine-PET was 87%, the specificity 75%, the positive predictive value 96% and the negative predictive value 43%. In all tumourous cases with positive tracer uptake the borderline area of the tumour was better defined by (11)C methionine-PET than by MRI. A positive (11)C methionine-PET is highly suspicious of a tumour, a negative one does not exclude it. (11)C methionine-PET seems to be more sensitive than MRI for differentiating between tumour and edema or gliosis. Simultaneous integration MRI and (11)C methionine-PET into cranial neuronavigation can facilitate cross total tumour removal in glioma surgery.