Noninvasive measurement of cardiovascular function in mice with high-temporal-resolution small-animal PET

Abstract

The aim of this study was to explore the feasibility of determining parameters of cardiovascular function in mice noninvasively by high-temporal-resolution imaging with a dedicated small-animal PET system.

Methods: Twenty-five anesthetized mice (28.8 +/- 4.6 g) were injected via an intravenous catheter with a 30-microL bolus of (18)F-FDG (8-44 MBq). The first 9 s of data were reconstructed into 30 frames of 0.3 s using filtered backprojection. The time-activity curve derived from a left ventricle volume of interest was corrected for tracer recirculation and partial volume. Cardiac output was calculated by the Stewart-Hamilton method, in which cardiac output is total injected activity divided by the area under the left ventricle time-activity curve. Cardiac output divided by body weight was defined as cardiac index; cardiac output divided by heart rate yielded the stroke volume. In 5 mice, measurements were repeated 2-4 times to assess reproducibility. In 4 mice, the hemodynamic response to dobutamine was examined by measuring heart rate, cardiac output, and stroke volume.

Results: The cardiac output averaged 20.4 +/- 3.4 mL/min; in the repeated measurements, the parameter displayed a mean percentage SD per mouse of 10% +/- 6%. The cardiac index averaged 0.73 +/- 0.19 mL/min/g and the stroke volume 45.0 +/- 6.9 microL, and both correlated with heart rate (r = 0.53, P = 0.007, and r = 0.49, P = 0.01, respectively). During dobutamine stress, heart rate increased from 423 +/- 50 to 603 +/- 30 beats/min (P = 0.002) and cardiac output increased from 18.5 +/- 1.9 to 32.0 +/- 4.2 mL/min (P = 0.008).

Conclusion: Parameters of cardiovascular function can be measured in mice noninvasively by radionuclide angiography using high-temporal-resolution small-animal PET. Measured values of cardiac output and stroke volume are reproducible and comparable to those obtained with MRI. The approach permits the monitoring of changes in cardiovascular function in response to pharmacologic intervention.

FIGURE 1

Example of LV time–activity curve and exponential fit correcting for recirculation. Area under upslope of LV time–activity curve and exponentially fitted downslope is finite.

FIGURE 2

Consecutive 0.3-s frames show passage of tracer bolus through RV cavity, lungs, and LV chamber of mouse on coronal and transverse slices. Times are those after start of image acquisition. For better anatomic orientation, PET scan is overlaid with coregistered CT scan.

FIGURE 3

Time–activity curves derived for VOIs placed in RV and LV chamber and over whole body (total count on secondary y-axis).

FIGURE 4

Correlations between cardiac output, cardiac index, and heart rate.

FIGURE 5

Results of repeated measurements of heart rate (A), cardiac output (B), and stroke volume (C) in 5 mice.

FIGURE 6

Individual changes in heart rate (A), cardiac output (B), and stroke volume (C) from baseline to dobutamine stress in 4 mice.