Measurement of cerebral astrocytic function using N-13 ammonia PET

Background and aims: N-13 ammonia (NH3) diffuses into cerebral tissue and is rapidly trapped in amino acid pools via glutamine synthetase. Anatomical localization of the enzyme (exclusively in astrocytes) and slow turnover of the amino acids pool allow measurement of cerebral astrocytic function as a retention of NH3. The purpose of this study was to investigate astrocytic function using NH3 PET in patients with stenotic lesion(s) in internal carotid arteries. Methods: Ten patients (2 women, 8 men, mean age 71 ± 3 years old) underwent F-18 FDG and N-13 NH3 PET. All had stenotic lesion(s) in one of the internal carotid arteries. Four of them underwent percutaneous intervention immediately after the PET scan, and repeated a PET study after a 3 months period to investigate the effect of the intervention. Retention of NH3 was investigated using a parametric image of standard uptake vale (SUV) anatomically coregistered to his/her MRI and FDG-PET images. Regions of interest were placed on the normal grey matter territories of middle cerebral arteries on right and left sides at the slice of centrum semiovale. SUV between pathologic and normal sides were compared using a t-test. Results: In normal grey and white matters, retention of NH3 was visually in parallel with FDG uptake except for hippocampus, pons and pituitary gland. These 3 regions showed increased retention of NH3 (compared to FDG) in all patients. Statistical analysis showed significantly lower retention of NH3 in apparently normal grey matter of the pathologic side (2.35±0.69) than that of normal side (2.53±0.6, p < 0.05 ). Percutaneous intervention increased regional NH3 SUV in 3/4 of the patients. Among 5 patients with subacute cortical infarction, 2 had infarcted cortices showing complete defect of FDG, while the remaining 3 had infarcted cortices showing preserved (normal to reduced) FDG. The former cortices showed minimal retention of NH3 (SUV: 0.18 and 0.58) while the latter cortices showed higher retention of NH3 (SUV: 2.45, 2.71 and 3.93, respectively) than that of the normal side (SUV: 2.23, 2.01 and 2.36, respectively) in each patient (Figure). Conclusions: Reduced retention of NH3 in the pathologic side may suggest impaired astrocytic function in apparently normal grey matter. Percutaneous intervention increased regional retention of NH3 in 3/4 patients, suggesting the improved astrocytic function. In conclusion, retention of NH3 can be a clinical marker of cerebral astrocytic function.