Quantification of regional cerebral blood flow in rats using an arteriovenous shunt and micro-PET

Measurement of regional cerebral blood flow (rCBF) in rodents can provide knowledge of pathophysiology of the cerebral circulation, but generally requires blood sampling for analysis during positron emission tomography (PET). We therefore tested the feasibility of using an arteriovenous (AV) shunt in rats for less invasive blood analysis.

Six anesthetized rats received [15O]H2O and [15O]CO PET scans with their femoral artery and vein connected by an AV shunt, the activity within which was measured with a germanium ortho-oxysilicate scintillation detector. The [15O]H2O was intravenously injected either at a faster or slower injection rate, while animals were placed either with their head or heart centered in the gantry. The time-activity curve (TAC) from the AV shunt was compared with that from the cardiac ventricle in PET image. The rCBF values were calculated by a nonlinear least-square method using the dispersion-corrected AV-shunt TAC as an input.

The AV-shunt TAC had higher signal-to-noise ratio, but also had delay and dispersion compared with the image-derived TAC. The delay time between the AV-shunt TAC and image-based TAC ranged from 11 to 21 s, while the dispersion was estimated to be ∼5 s as a time constant of the dispersion model of exponential function, and both were properly corrected. In a steady-state condition of [15O]CO PET, the blood activity concentration by AV-shunt TAC was also comparable in height with the image-based TAC corrected for partial volume. Whole-brain CBF values measured by [15O]H2O were 0.37±0.04 (mean±S.D.) ml/g/min, partition coefficient was 0.73±0.04 ml/g, and the CBF varied in a linear relationship with partial pressure of carbon dioxide during each scan.

The AV-shunt technique allows less invasive, quantitative and reproducible measurement of rCBF in [15O]H2O PET studies in rats than direct blood sampling and radioassay.