In vivo quantitation of glucose metabolism in mice using small-animal PET and a microfluidic device

The challenge of sampling blood from small animals has hampered the realization of quantitative small-animal PET. Difficulties associated with the conventional blood-sampling procedure need to be overcome to facilitate the full use of this technique in mice. Methods: We developed an automated blood-sampling device on an integrated microfluidic platform to withdraw small blood samples from mice. We demonstrate the feasibility of performing quantitative small-animal PET studies using F-18-FDG and input functions derived from the blood samples taken by the new device. F-18-FDG kinetics in the mouse brain and myocardial tissues were analyzed. Results: The studies showed that small (similar to 220 nL) blood samples can be taken accurately in volume and precisely in time from the mouse without direct user intervention. The total blood loss in the animal was < 0.5% of the body weight, and radiation exposure to the investigators was minimized. Good model fittings to the brain and the myocardial tissue time-activity curves were obtained when the input functions were derived from the 18 serial blood samples. The R-2 values of the curve fittings are > 0.90 using a F-18-FDG 3-compartment model and > 0.99 for Patlak analysis. The F-18-FDG rate constants K-1*, k(2)*, k(3)*, and k(4),*,obtained for the 4 mouse brains, were comparable. The cerebral glucose metabolic rates obtained from 4 normoglycemic mice were 21.5 +/- 4.3 mu mol/min/100 g (mean +/- SD) under the influence of 1.5% isoflurane. By generating the whole-body parametric images of K*(FDG) (mL/min/g), the uptake constant of F-18-FDG, we obtained similar pixel values as those obtained from the conventional regional analysis using tissue time-activity curves. Conclusion: With an automated microfluiclic blood-sampling device, our studies showed that quantitative small-animal PET can be performed in mice routinely, reliably, and safely in a small-animal PET facility.