PT - JOURNAL ARTICLE AU - Stephanie L. Thorn AU - Robert A. deKemp AU - Tyler Dumouchel AU - Ran Klein AU - Jennifer M. Renaud AU - R. Glenn Wells AU - Michael H. Gollob AU - Rob S. Beanlands AU - Jean N. DaSilva TI - Repeatable Noninvasive Measurement of Mouse Myocardial Glucose Uptake with <sup>18</sup>F-FDG: Evaluation of Tracer Kinetics in a Type 1 Diabetes Model AID - 10.2967/jnumed.112.110114 DP - 2013 Sep 01 TA - Journal of Nuclear Medicine PG - 1637--1644 VI - 54 IP - 9 4099 - http://jnm.snmjournals.org/content/54/9/1637.short 4100 - http://jnm.snmjournals.org/content/54/9/1637.full SO - J Nucl Med2013 Sep 01; 54 AB - A noninvasive and repeatable method for assessing mouse myocardial glucose uptake with 18F-FDG PET and Patlak kinetic analysis was systematically assessed using the vena cava image–derived blood input function (IDIF). Methods: Contrast CT and computer modeling was used to determine the vena cava recovery coefficient. Vena cava IDIF (n = 7) was compared with the left ventricular cavity IDIF, with blood and liver activity measured ex vivo at 60 min. The test–retest repeatability (n = 9) of Patlak influx constant Ki at 10–40 min was assessed quantitatively using Bland–Altman analysis. Myocardial glucose uptake rates (rMGU) using the vena cava IDIF were calculated at baseline (n = 8), after induction of type 1 diabetes (streptozotocin [50 mg/kg] intraperitoneally, 5 d), and after acute insulin stimulation (0.08 mU/kg of body weight intraperitoneally). These changes were analyzed with a standardized uptake value calculation at 20 and 40 min after injection to correlate to the Patlak time interval. Results: The proximal mouse vena cava diameter was 2.54 ± 0.30 mm. The estimated recovery coefficient, calculated using nonlinear image reconstruction, decreased from 0.76 initially (time 0 to peak activity) to 0.61 for the duration of the scan. There was a 17% difference in the image-derived vena cava blood activity at 60 min, compared with the ex vivo blood activity measured in the γ-counter. The coefficient of variability for Patlak Ki values between mice was found to be 23% with the proposed method, compared with 51% when using the left ventricular cavity IDIF (P &lt; 0.05). No significant bias in Ki was found between repeated scans with a coefficient of repeatability of 0.16 mL/min/g. Calculated rMGU values were reduced by 60% in type 1 diabetic mice from baseline scans (P &lt; 0.03, ANOVA), with a subsequent increase of 40% to a level not significantly different from baseline after acute insulin treatment. These results were confirmed with a standardized uptake value measured at 20 and 40 min. Conclusion: The mouse vena cava IDIF provides repeatable assessment of the blood time–activity curve for Patlak kinetic modeling of rMGU. An expected significant reduction in myocardial glucose uptake was demonstrated in a type 1 diabetic mouse model, with significant recovery after acute insulin treatment, using a mouse vena cava IDIF approach.