Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2452061973
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Movie and Figure
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chin, B. B.
Right arrow Articles by Jaszczak, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chin, B. B.
Right arrow Articles by Jaszczak, R. J.
(Radiology 2007;245:440-448.)
© RSNA, 2007

Experimental Studies

Left Ventricular Functional Assessment in Mice: Feasibility of High Spatial and Temporal Resolution ECG-gated Blood Pool SPECT1

Bennett B. Chin, MD, Scott D. Metzler, PhD, Anthony Lemaire, MD, Antonio Curcio, MD, Sreekanth Vemulapalli, BS, Kim L. Greer, CNMT, Neil A. Petry, MS, Timothy G. Turkington, PhD, R. Edward Coleman, MD, Howard Rockman, MD, and Ronald J. Jaszczak, PhD

1 From the Departments of Radiology (B.B.C., K.L.G., N.A.P., T.G.T., R.E.C., R.J.J.) and Medicine (A.L., A.C., S.V., H.R.), Duke University School of Medicine, Box 3808, Durham, NC 27710; and Department of Radiology, University of Pennsylvania, Philadelphia, Pa (S.D.M.). Received November 20, 2006; revision requested January 22, 2007; revision received February 23; final version accepted April 2. Address correspondence to B.B.C. (e-mail: chin0004{at}mc.duke.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 
Purpose: To prospectively determine feasibility of evaluating murine left ventricular (LV) function with electrocardiographically (ECG)-gated blood pool single photon emission computed tomography (SPECT).

Materials and Methods: All animal studies had institutional animal care and use committee approval. SPECT was performed with conventional time-binned acquisition (eight frames per ECG cycle) in normal mice (normal group A, n = 6) and mice with myocardial infarction (MI) (n = 8). To determine feasibility of high temporal resolution and rapid data acquisition, another group of normal mice (normal group B, n = 4) underwent imaging with conventional (eight-frame) time-binned and list-mode (LM) acquisitions. LM acquisitions were reconstructed with eight and 16 frames per ECG cycle and 10 minutes of data (short LM). SPECT images were assessed visually, and LV-to–lung background activity ratios were calculated. LV end-systolic and end-diastolic volumes were defined with a phase analysis and threshold method. LV ejection fraction (LVEF) was calculated from LV volumes and count-based methods (n = 18 mice). Fractional shortening (FS) at echocardiography defined MI dysfunction (mild MI: FS ≥ 50%; severe MI: FS < 50%). Group means were compared for significant differences with analysis of variance.

Results: ECG-gated blood pool SPECT demonstrated normal, concentric LV contraction in all normal mice (n = 10). LV-to–lung background ratio was more than 10:1 (range, 10.3–29.4; n = 18). Focal wall motion abnormalities were detected at SPECT both visually and with phase analysis in all mice with severe MI (n = 5). Mice with severe MI had significantly lower LVEF than normal group A mice (32% ± 14 [standard deviation] vs 64% ± 8%; P < .001). All mice with mild MI (n = 3) had normal contraction and LVEF. In paired acquisitions in normal group B mice, all reconstructions (n = 16) showed normal LV contraction. LVEF was not significantly different (P = .88) between time-binned (71% ± 12), eight-frame LM (71% ± 12), 16-frame LM (77% ± 10), and short LM (73% ± 14) reconstructions.

Conclusion: Murine LV functional assessment is feasible with high spatial and temporal resolution ECG-gated blood pool SPECT. LV dysfunction can be quantified and focal wall motion abnormalities detected in the MI model of heart failure.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/245/2/440/DC1

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 
The wide variety and availability of genetically and surgically created murine models of heart failure provide a method for interrogating the complex molecular mechanisms of heart failure. Advanced genomics and transgenic mice have already greatly facilitated the elucidation of several molecular signaling pathways in heart failure (1,2). To relate these molecular mechanisms to a functional outcome is a primary goal because functional measurements are clinically useful in characterizing, monitoring, and predicting outcomes in human studies.

Two-dimensional echocardiography for left ventricular (LV) assessment in mice is widely available, rapid, and technically easy to perform. This technique, however, may be less accurate or reproducible compared with three-dimensional (3D) techniques when measuring indexes of function such as LV ejection fraction (LVEF) or ventricular volumes. Magnetic resonance (MR) imaging and high spatial and temporal resolution computed tomography (CT) have shown good accuracy in the evaluation of LV function and morphology (3–10). More recently, 3D echocardiography has also been reported to have high correlation with MR imaging findings (11). These techniques, however, are currently investigational and may require highly specialized instrumentation and trained personnel—factors that may not be readily available. Technical or logistic factors, limited availability, or high cost of dedicated small-animal MR imaging or CT may also limit the feasibility for widespread use.

Recently developed high-spatial-resolution single photon emission CT (SPECT) radionuclide techniques permit submillimeter resolution (12–15). Combining high spatial resolution with high temporal resolution, electrocardiographically (ECG)-gated SPECT imaging may be an alternative method for evaluating LV function in murine models of heart failure (16). In addition, radionuclide techniques are quantitative and may be highly reproducible, as well as rapidly performed. Thus, the purpose of our study was to prospectively determine the feasibility of evaluating murine LV function with ECG-gated blood pool SPECT.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 
Phantom Studies
Small hot rod phantom.—To demonstrate the capability to obtain submillimeter resolution with this system, hot rod phantom SPECT imaging was performed (K.L.G., R.J.J.). This phantom (Ultra-Micro Hot Spot; Data Spectrum, Hillsborough, NC) has hollow channels of 0.75, 1.0, 1.35, 1.7, 2.0, and 2.4 mm in diameter and center-to-center spacing of two times the diameter of the respective hollow channels. Acquisition and reconstruction parameters included the following: use of 5 µCi/µL (0.185 MBq/µL) technetium 99m (99mTc) pertechnetate, 30 seconds per view, 120 views, 360° rotation, 256 x 128 matrix, 3.1-cm radius of rotation, 1.0-mm tungsten pinhole (1), ordered-subset expectation maximization, iterative reconstruction (four subsets, five iterations), and a reconstructed voxel size of 0.125 µL.

Small hollow-sphere phantom.—Small spheres with volumes of 31, 61, 125, and 250 µL and diameters of 3.95, 4.95, 6.23, and 7.86 mm, respectively (Micro-volume Hollow Spheres; Data Spectrum), were filled with 6 µCi/µL (0.222 MBq/µL) 99mTc (B.B.C., K.L.G., R.J.J.). SPECT imaging with 1.0-mm pinhole collimation was performed with parameters adjusted to simulate the time and activity used for the mouse blood pool acquisitions described below. The accuracy of the sphere volumes was verified by weighing the spheres with a high-precision scale before and after they were filled with water. A threshold value was determined by matching the actual sphere volumes with cawith calculated volumes that were based on a percentage of the peak activity within the spheres. This threshold was applied to the animal data to determine if the phantom data results accurately modeled animal data.

Animal Studies
All animal studies were approved by the Duke University Institutional Animal Care and Use Committee. Protocols were developed and performed under American Veterinary Association guidelines at our Association for Assessment and Accreditation of Laboratory Animal Care–certified facility. To assess ECG-gated blood pool SPECT in this setting, wild-type C57Bl/6 normal control mice (normal group A, n = 6) were compared with wild-type C57Bl/6 mice with myocardial infarction (MI) (n = 8) (B.B.C., A.C., K.L.G., R.J.J., S.D.M., A.L., S.V.). All mice used in our studies were adult mice ranging in age between 12 and 20 weeks and weighing an average of 28.0 g (range, 20.8–38.5 g). Prior to induction of MI, mice were anesthetized with a mixture of ketamine (100 mg per kilogram of body weight; intraperitoneal injection), and xylazine (2.5 mg/kg; intraperitoneal injection) (A.C., A.L., H.R.). Under a dissecting microscope (Nissho Optical TZ-240; LABTEK, Campbell, Calif), animals were placed in the supine position, and a midline cervical incision was made to expose the trachea. After endotracheal intubation, the cannula was connected to a volume-cycled rodent ventilator (Model 683; Harvard Apparatus, Holliston, Mass) using room air with a stroke volume of 0.25 mL and a respiratory rate of 100 respirations per minute. The chest cavity was entered in the fourth intercostal space at the left sternal border through a small incision, and MI was induced by ligating the left anterior descending coronary artery with 8.0 Prolene suture (Ethicon, Somerville, NJ) at the site of the vessel's emergence past the tip of the left atrium. Extent of MI, however, was variable because of the normal variation in murine coronary anatomy. Animals recovered for 8 weeks before undergoing radionuclide imaging. A second set of normal control mice (normal group B; age range, 12–16 weeks; mean weight, 30.2 g ± 1.5 [standard deviation]; n = 4) of the same approximate age and strain were imaged to determine the feasibility of rapid and high-temporal-resolution data acquisition with paired conventional time-binned and list-mode acquisitions, as described below.

Red Blood Cell Labeling
The in vivo red blood cell labeling procedure with 99mTc pertechnetate is similar to the procedure used in humans. The in vivo 99mTc red blood cell labeling efficiency is more than 95% at 1 hour after the administration of 99mTc pertechnetate by intravenous, retroorbital sinus, or subcutaneous labeling methods (17). Briefly, 15 µg/kg (0.3–0.4 µg for 20–30-g mice) stannous pyrophosphate (Cis-Pyro; Cis, Bedford, Mass) was injected intravenously through the tail vein (B.B.C., N.P., S.V.). Approximately 15–20 minutes later, 12.0 mCi (444 MBq) 99mTc pertechnetate was injected intravenously in the opposite tail vein.

SPECT Imaging
Image acquisition.—For all mice (normal group A [n = 6], normal group B [n = 4], and MI [n = 8]), anesthesia was induced with 1.0%–2.0% isofluorane through a nose cone (EZ anesthesia 1000; Euthanex, Palmer, Pa) (B.B.C., S.V., S.D.M., K.L.G., A.C., A.L.). ECG-gated data were acquired in eight time bins by using a specially modified ECG gating system capable of accurate gating at 600 beats per minute (AccuSync 71; AccuSync Medical Research, Milford, Conn), and all images were acquired at heart rates of less than 500 beats per minute. Pediatric ECG leads (Blue Sensor; Ambu, Ballerup, Denmark) were used. Gated SPECT imaging was performed with a triple-detector gamma camera (Triad Trionix-XLT; Research Laboratories, Twinsburg, Ohio) as previously described (15) (Fig 1). Image acquisition began 60 minutes after the 99mTc pertechnetate injection. Three custom-designed pyramidal collimators (focal length, 20.0 cm) with 1.0-mm tungsten apertures simultaneously acquired projection data with the following parameters: 45–60 heartbeats per projection (approximately 30–45 minutes, depending on heart rate), a matrix of 128 x 64, 2° per projection, 360° rotation (180 projections per head), step and shoot, circular orbit, and a radius of rotation of 2.7–3.1 cm. This corresponds to an image magnification factor of 7.4 for a radius of 2.7 cm. The imaging field of view included the entire cardiac blood pool and adjacent thorax.


Figure 1
View larger version (172K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: SPECT instrumentation. A, Typical 25-g mouse. B, Mouse imaging with high magnification and radius of rotation of 27 mm. C, Three-detector gamma camera with 1.0-mm tungsten pinholes mounted to pyramidal collimators.

 
To determine the capability of performing very high temporal resolution imaging and rapid data acquisition, a paired set of conventional time-binned acquisitions and list-mode data acquisitions were performed in a separate group of normal control mice of same age and strain (normal group B, n = 4) (B.B.C., S.V., S.D.M., K.L.G.). In two mice, list-mode acquisition was performed first, followed by time-binned acquisition; in the other two mice, the order was reversed. In two mice, a single pinhole collimator and detector were used; thus, the count acquisition for these studies was approximately one-third of that for the other studies. The other collimator data were used in a collimator comparison study, results of which will be reported separately.

List-mode acquisition stores information of each detected photon event sequentially with timing and gating markers. Data were retrospectively separated into eight and 16 time bins by evenly partitioning the photons in each cardiac cycle into the eight and 16 time bins. To demonstrate the potential for rapid data acquisition, short list-mode SPECT data sets were reconstructed from the equivalent of the first 10 minutes of data acquisition by using eight time frames (S.D.M.).

Image reconstruction.—All gated data were reconstructed with an iterative algorithm that incorporates collimator response function and separate high-precision calibrations to account for individual variations in detector positions (15,18–20). Reconstruction of data was performed for four subsets and 10 iterations. The fifth iteration was chosen for data analysis and display on the basis of a subjective visual assessment that considered the compromise between image contrast and noise. The reconstructed voxel size of 0.5 mm3 corresponds to a volume of 0.125 µL/voxel. Data were visualized and analyzed with a flexible data analysis program (SPECTER 4.0) written in our laboratory, and visualizations of surface rendering were displayed with a commercially available workstation (Xeleris; GE Medical Systems, Waukesha, Wis) by using standard commercially available cardiac software (Cedars-Sinai QGS-QPS; Los Angeles, Calif) (B.B.C.).

Data Analysis
Visual assessment.—All reconstructed SPECT images were inspected in cine mode in the transverse, coronal, and sagittal planes at the mid-LV level. Images were visually assessed for delineation of the right ventricular and LV blood pool boundaries and were subjectively judged as excellent (clear delineation without ambiguity at end diastole [ie, activity in the septum of approximately 50% of the peak LV activity]), adequate (clear delineation with definite separation at end systole but not at end diastole), or inadequate (ambiguous ventricular separation) by a single observer (B.B.C., with more than 15 years of experience in cardiac SPECT imaging). Wall motion abnormalities were also assessed visually for concentric contraction (normal wall motion) or focal wall motion abnormalities by the same observer.

Quantitative assessment.—The LV blood pool–to–lung background ratio was assessed by computing the mean activity in 16-voxel-square regions of interest in the center of the LV and in the right lung above the level of the aortic arch on the transverse end-diastolic image. The LV end-diastolic and end-systolic frames were first selected by determining the time frames with the maximum and minimum LV counts, respectively (B.B.C.). Peak activity was defined as the maximum voxel value within the LV at end diastole. Boundary definition of ventricular cavities was performed by combining information from Fourier analysis and a count-based thresholding technique (B.B.C., S.D.M.). To define atrial and ventricular separation, the Fourier time coefficients of each voxel were determined from the four-dimensional data set. Each voxel Vi was parametrically represented as Vi(Ø) = Ai + Bisin(Ø) + Cicos (Ø), where Ø is the cardiac phase (0 ≤ Ø < 2{pi}) and the average ungated value of Vi is Ai. The Fourier coefficients Ci are positive in the ventricles because their Vi(0) is greater than Ai and have opposite phase in the atria because their Vi(0) is less than Ai. Consequently, the 3D map of Ci values was used to define the chamber boundaries. Bi is the Fourier coefficient for the sine component; this does not have a direct physiologically defined relevance but is a mathematical component of the cyclic activity over time. This process is analogous to using a 3D phase image to define the separation of atria and ventricles, as previously described (21). These regions, with their precise locations, were superimposed on the coregistered end-diastolic and end-systolic data sets for additional processing with a threshold technique (B.B.C., T.G.T.).

To determine a threshold method for LV boundary detection used in conjunction with phase analysis, the small hollow-sphere phantom threshold values described above were compared by using an empiric threshold method (B.B.C.). The empiric method used a threshold of 50% peak activity for end diastole and a lower threshold value (42% peak activity) for end systole. The rationale for choosing a 50% peak threshold for end diastole was based on small partial-volume effects for LV end-diastolic diameters, which are greater than twice the full width at half maximum of the system's spatial resolution. For our system at a radius of rotation of 2.8 cm, the resolution for 99mTc is less than 1.0 mm. By using three normal group A mice, a value of 42% was determined to best match the reported end-systolic volumes (ESVs) at MR imaging of mice of the same age and strain (8). Because the partial-volume effect is greater at smaller ESV values, this lower threshold partially compensates for this effect.

In addition to defining LVEF by using voxel volumes, we also defined LVEF by using the total counts of the voxels at end diastole and end systole. With both methods, the LVEF was defined as [(value at end diastole – value at end systole)/value at end diastole]·100% (B.B.C.). All values reported for ventricular volume and LVEF measurements were calculated by using the volume measurement method unless otherwise stated.

Transthoracic Echocardiography
Echocardiography was performed in conscious mice from all groups (Vevo 660 system, 30-mHz transducer; Visualsonics, Toronto, Ontario, Canada) within 1 week of the radionuclide study, as previously described (A.C., A.L., H.R.) (22).Two-dimensional echocardiography was used to determine short-axis midventricular positioning, and M-mode imaging was used for LV measurements. Three separate echocardiography measurements were averaged, and the mean was used for comparison with SPECT data. Percentage fractional shortening was defined as [(LVDed – LVDes)/LVDed]·100%, where LVDed is the LV end-diastolic diameter and LVDes is the LV end-systolic diameter in millimeters on the midventricular short-axis view (23). Cardiac dysfunction after MI was empirically classified on the basis of fractional shortening as mild (fractional shortening, ≥50%), or severe (fractional shortening, <50%).

Statistical Analysis
All results are expressed as means ± 1 standard deviation. Analysis of variance was used to test for significant differences in LVEF and LV EDVs and ESVs in the normal groups with conventional time-binned mode and list-mode reconstructions (B.B.C.). In the instance where several comparisons were made between the same groups, a multiple-comparisons adjustment was made when significance was tested. Thus, P < .008 (.05/6) was considered to indicate a statistically significant difference for six comparisons. Linear regression was used to assess the relationship between LVEF calculated from volumes and LVEF calculated from counts. The data were tested for significance with standard statistical analysis software (JMP, version 6.0; SAS Institute, Cary, NC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 
Phantom Studies
The Ultra Micro Hot Spot phantom demonstrated submillimeter spatial resolution with the 1.0-mm aperture pinhole (Fig 2). The Micro-volume Hollow Spheres demonstrated high image contrast; however, the thresholds (33%–35% maximal activity) that produced accurate phantom volumes resulted in markedly enlarged blood pool volumes that did not sufficiently represent the dynamic physiologic volumes in mice. The empirically derived thresholds—50% for end diastole and 42% for end systole—provided better visual concordance with boundary definition, and this method was used for all subsequent volume and LVEF calculations.


Figure 2
View larger version (62K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Hot rod phantom (left) and transverse SPECT image (right). The smallest (0.75-mm-diameter) channels (right arrow) can be identified, demonstrating submillimeter spatial resolution capability. Left arrow indicates smallest channel diameter (0.75 mm); channel diameters progress clockwise from 1.0 to 1.35, 1.7, 2.0, and 2.4 mm.

 
Animal Studies
The overall image quality was excellent, with clear delineation of the ventricular blood pool in all normal group A mice (n = 6), normal group B mice (n = 4), and mice with MI (n = 8). When data were viewed in cine format, all normal group A mice showed concentric LV contraction and normal LV wall motion. The LV size appeared proportionally normal compared with the size of the right ventricle. The parametric images were concordantly normal in all normal group A mice, with uniform phase within the LV cavity and at the apex (Tables 1, 2; Fig 3; Movie [http://radiology.rsnajnls.org/cgi/content/full/245/2/440/DC1]).


View this table:
[in this window]
[in a new window]

  		Table 1. SPECT and Echocardiography Data

 

View this table:
[in this window]
[in a new window]

  		Table 2. LV Measurements in Mouse Groups

 

Figure 3
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: SPECT images in normal control group A mouse 4 (Table 1) (LVEF = 72%) at end diastole and end systole. LV contraction is concentric, and ESV is small. Top and second rows show short-axis sections from apex to base (left to right) at end diastole and end systole, respectively. Third and fourth rows show corresponding vertical long-axis sections from septum to lateral wall (left to right) at end diastole and end systole, respectively. Fifth and sixth rows show corresponding horizontal long-axis sections from septum to lateral wall (left to right) at end diastole and end systole, respectively. Top row of Movie (http://radiology.rsnajnls.org/cgi/content/full/245/2/440/DC1) shows ventricular wall motion in cine format.

 
In the group of mice with MI, all mice with severe MI (n = 5) showed an apical wall motion abnormality at visual inspection (Fig 4) and also at phase analysis (Fig 5). The LV cavity in all five mice was visually dilated compared with the right ventricular cavity. The right ventricle, however, demonstrated preservation of normal ventricular size and wall motion at visual inspection. SPECT quantification showed that the largest LV ESVs and EDVs corresponded to the largest end-systolic and end-diastolic diameters at echocardiography, respectively (see data for mouse 4 in severe MI group in Table 1). The LVEF was significantly lower in mice with severe MI than in normal group A mice (32% ± 14 vs 64% ± 8; P < .001).


Figure 4
View larger version (52K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: SPECT images in severe MI group mouse 5 (Table 1) (LVEF at SPECT = 41%). Note the LV apical hypokinesis at end systole (long arrow in second row; short arrows in fourth and sixth rows) and relatively well-preserved contraction toward the base (short arrow in second row; long arrows in fourth and sixth rows). Orientation is the same as in Figure 3. Top and second rows show end-diastolic and end-systolic short-axis sections from apex to base (left to right), respectively. Third and fourth rows show end-diastolic and end-systolic vertical long-axis sections from septum to lateral wall (left to right), respectively. Fifth and sixth rows show end-diastolic and end-systolic horizontal long-axis sections from septum to lateral wall (left to right), respectively. Bottom row of Movie (http://radiology.rsnajnls.org/cgi/content/full/245/2/440/DC1) shows ventricular wall motion in cine format.

 

Figure 5
View larger version (57K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Parametric images of amplitude and phase in normal control group A mouse 4 (top row) and severe MI group mouse 5 (same mice as in Figs 3 Figs 4, respectively). On phase images, gray scale is adjusted to show ventricular contraction as high or "bright" and atrial contraction as low or "dark." In MI group mouse, there is apical dilation (white arrow) and abnormal apical hypokinesis (black arrow) that is out of phase with normal LV contraction.

 
In the group of mice with mild MI, SPECT showed normal LV wall motion, according to both visual inspection and phase analysis, in all mice (n = 3). In this group, all phase images of the LV were normal and indistinguishable from those in normal group A. The LV sizes at end diastole and end systole were not significantly different in the mild MI group from those in normal group A (P = .18 and P = .74, respectively). Overall, the LVEF in the mild MI group was not significantly different (P = .16) from that in normal group A.

In the second group of normal control mice (normal group B, n = 4), there was excellent visualization of the LV blood pool, normal concentric LV contraction, and normal LV wall motion for all reconstructions (n = 16). The LV-to–lung background ratio was high for all reconstructed studies (20.9 ± 4.7, n = 16). All list-mode acquisitions and reconstructions were visually similar to the time-binned acquisitions (Fig E1, http://radiology.rsnajnls.org/cgi/content/full/245/2/440/DC1). The short 10-minute acquisitions appeared to have lower counts but demonstrated clear separation of the left and right ventricular boundaries. The SPECT measurements of LVEF, LV EDV, and LV ESV were not statistically different between these groups (P = .88, P = .94, and P = .92, respectively) (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 3. Paired SPECT Measurements Acquired with Different Acquisition and Reconstruction Methods in Four Normal Control (Group B) Mice

 
Overall, the LV-to–lung background ratio for all groups (normal group A [n = 6], normal group B [n = 4], and MI [n = 8]) and all reconstructions (including the four reconstructions performed for all normal group B mice) was greater than 10:1 (range, 10.3–29.4; n = 30). For all acquisitions and reconstructions described above, LVEF at SPECT calculated from the total counts within the voxels at end diastole and end systole showed high correlation (R2 = 0.99; n = 30) with LVEF calculated from the volume measurements (Fig 6).


Figure 6
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Graph demonstrates that LVEF calculated from total counts shows high correlation with values calculated from voxel volumes over a wide range of values for all mice and all reconstructions (n = 30).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 
Our study results demonstrate the ability to perform both high spatial and temporal resolution ECG-gated blood pool SPECT to visualize and quantify LV function in both normal mice and mice with MI. ECG-gated blood pool SPECT can also depict and objectively display a focal wall motion abnormality after MI. SPECT LV volumes and LVEF measurements were comparable with those previously reported at MR imaging. Results with our gated list-mode technique demonstrate the capability to perform high-temporal-resolution SPECT imaging and also show the feasibility of rapid data acquisition with preservation of excellent image quality and high contrast.

Our calculated mean LVEF at SPECT (64% ± 8) was comparable with values previously reported. In MR imaging measurements of normal mice of the same strain and age as ours, the mean LV EDV was reported to be 63.6 µL ± 6.6, the mean ESV was 23.5 µL ± 4.4, and the mean LVEF was 64.6% ± 3.5 (8). Reported LVEF of normal control mice at 3D echocardiography (63% ± 5) was also comparable (11). Finally, ECG-gated SPECT of myocardial perfusion with 99mTc tetrofosmin has also been performed (14,24) and yielded a comparable LVEF of 60% ± 9 in normal mice (14).

The use of 99mTc ECG-gated blood pool SPECT to assess ventricular function has a number of advantages. First, in the MI model of heart failure, the use of blood pool imaging avoids the issue of variability in boundary detection when perfusion is low or absent. Second, blood pool imaging allows the simultaneous evaluation of both LV and right ventricular size and function; this may be particularly useful in models of right ventricular failure (25,26). In human studies (21,27), ECG-gated blood pool SPECT for evaluation of right ventricular ejection fraction has demonstrated excellent correlation with established techniques. Third, ECG-gated blood pool SPECT has demonstrated excellent reproducibility in a larger rat model, providing evidence supporting the possibility of high reproducibility with this technique (28). Fourth, small-animal SPECT systems currently have finer spatial resolution than the majority of state-of-the-art dedicated animal positron emission tomography (PET) scanners (29). Finally, this ECG-gated blood pool SPECT method is technically feasible without the need for a dedicated small-animal scanner or dedicated PET radiotracer synthesis (15).

Our gated list-mode acquisition method produced high temporal and spatial resolution studies. Reconstructions with even higher temporal and spatial resolution are possible; however, these reconstructions also have an adverse effect on data set size and reconstruction speed. The 10-minute data acquisition showed high contrast and excellent image quality; this demonstrates the feasibility of rapid acquisition studies.

A limitation of our study was the correlation of ECG-gated blood pool SPECT results with those of two-dimensional echocardiography. M-mode echocardiography evaluates the midventricular short-axis section without specific attention to distal regional wall motion abnormalities. Accurate volumes could not be determined with M-mode echocardiography; thus, accurate calibration of ECG-gated blood pool SPECT could not be performed. Future studies involving 3D imaging techniques will be needed to calibrate the SPECT volumes and to better assess the accuracy of subsequent volume measurements and boundary detection algorithms. The complex issue of boundary detection has been the subject of extensive investigation in human studies (27), and developments are still ongoing. Likewise, algorithms developed specifically for small-animal cardiac imaging require further improvement. Higher temporal resolution, as demonstrated with the 16-frame-per-ECG-cycle list-mode data, may also improve the accuracy of ventricular measurements (30) and may correlate better with accepted standards (31); however, a 3D standard with high temporal resolution for comparison is technically difficult to achieve with other modalities such as MR imaging. Finally, our study lacked a sham operation in the normal control comparison groups. Focal apical wall motion abnormalities caused by thoracotomy, however, are unlikely and have not been reported.

In summary, high spatial and temporal resolution ECG-gated blood pool SPECT of murine LV function is feasible in both normal and post-MI mouse models. Feasibility of rapid data acquisition was demonstrated with list-mode acquisitions. The potential advantages of radionuclide techniques include quantitative accuracy and reproducibility, as well as the potential for wide availability, low cost, technical simplicity, and rapid data acquisition. Further studies are necessary to define the reproducibility and accuracy of this method.

Practical applications: ECG-gated blood pool SPECT is a high temporal and spatial resolution imaging modality for evaluating in vivo murine LV function. The ability to rapidly acquire high-contrast images also facilitates experimental efficiency. Further development could lead to a rapid, accurate, and highly reproducible method for evaluating heart failure therapies in mice. Successful preclinical heart failure therapies in mice, as evaluated with ECG-gated blood pool SPECT, may then provide evidence supporting the rationale for human clinical trials similarly intended to improve functional outcomes.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: ECG = electrocardiography • EDV = end-diastolic volume • ESV = end-systolic volume • LV = left ventricle • LVEF = LV ejection fraction • MI = myocardial infarction • 3D = three-dimensional

Guarantor of integrity of entire study, B.B.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, B.B.C.; experimental studies, B.B.C., S.D.M., A.L., A.C., S.V., K.L.G., N.A.P., T.G.T., H.R., R.J.J.; statistical analysis, B.B.C.; and manuscript editing, all authors

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
References
 

  1. Rockman HA, Koch WJ, Lefkowitz RJ. Seven-transmembrane-spanning receptors and heart function. Nature 2002;415:206–212. [CrossRef][Medline]
  2. Naga Prasad SV, Jayatilleke A, Madamanchi A, Rockman HA. Protein kinase activity of phosphoinositide 3-kinase regulates beta-adrenergic receptor endocytosis. Nat Cell Biol 2005;7:785–796. [CrossRef][Medline]
  3. Badea C, Hedlund LW, Johnson GA. Micro-CT with respiratory and cardiac gating. Med Phys 2004;31:3324–3329. [CrossRef][Medline]
  4. Ruff J, Wiesmann F, Hiller KH, et al. Magnetic resonance microimaging for noninvasive quantification of myocardial function and mass in the mouse. Magn Reson Med 1998;40:43–48. [Medline]
  5. Streif JU, Herold V, Szimtenings M, et al. In vivo time-resolved quantitative motion mapping of the murine myocardium with phase contrast MRI. Magn Reson Med 2003;49:315–321. [CrossRef][Medline]
  6. Wiesmann F, Szimtenings M, Frydrychowicz A, et al. High-resolution MRI with cardiac and respiratory gating allows for accurate in vivo atherosclerotic plaque visualization in the murine aortic arch. Magn Reson Med 2003;50:69–74. [CrossRef][Medline]
  7. Wiesmann F, Ruff J, Engelhardt S, et al. Dobutamine-stress magnetic resonance microimaging in mice: acute changes of cardiac geometry and function in normal and failing murine hearts. Circ Res 2001;88:563–569. [Abstract/Free Full Text]
  8. Wiesmann F, Ruff J, Hiller KH, Rommel E, Haase A, Neubauer S. Developmental changes of cardiac function and mass assessed with MRI in neonatal, juvenile, and adult mice. Am J Physiol Heart Circ Physiol 2000;278:H652–H657. [Abstract/Free Full Text]
  9. Zhou R, Pickup S, Glickson JD, Scott CH, Ferrari VA. Assessment of global and regional myocardial function in the mouse using cine and tagged MRI. Magn Reson Med 2003;49:760–764. [CrossRef][Medline]
  10. Siri FM, Jelicks LA, Leinwand LA, Gardin JM. Gated magnetic resonance imaging of normal and hypertrophied murine hearts. Am J Physiol 1997;272:H2394–H2402. [Medline]
  11. Dawson D, Lygate CA, Saunders J, et al. Quantitative 3-dimensional echocardiography for accurate and rapid cardiac phenotype characterization in mice. Circulation 2004;110:1632–1637. [Abstract/Free Full Text]
  12. Beekman FJ, van der HF, Vastenhouw B, et al. U-SPECT-I: a novel system for submillimeter-resolution tomography with radiolabeled molecules in mice. J Nucl Med 2005;46:1194–1200. [Abstract/Free Full Text]
  13. Beekman FJ, McElroy DP, Berger F, Gambhir SS, Hoffman EJ, Cherry SR. Towards in vivo nuclear microscopy: iodine-125 imaging in mice using micro-pinholes. Eur J Nucl Med Mol Imaging 2002;29:933–938. [CrossRef][Medline]
  14. Constantinescol A, Choquet P, Monassier L, Israel-Jost V, Mertz L, Elfertak L. Left ventricular perfusion and function in normal mice using pinhole gated single photon emission computed tomography [abstr]. Mol Imaging Biol 2005;7(P):214.
  15. Metzler SD, Jaszczak RJ, Patil NH, Vemulapalli S, Akabani G, Chin BB. Molecular imaging of small animals with a triple-head SPECT system using pinhole collimation. IEEE Trans Med Imaging 2005;24:853–862. [CrossRef][Medline]
  16. Chin BB, Lemaire A, Metzler SD, et al. ECG gated ultrahigh resolution pinhole single photon emission computed tomography for in vivo murine myocardial assessment [abstr]. Mol Imaging Biol 2005;7(P):176.
  17. Vemulapalli S, Metzler SD, Petry N, et al. In vivo murine 99mTc red blood cell labeling methods for cardiac function and vascular blood pool localization [abstr]. Mol Imaging Biol 2005;7(P):301.
  18. Metzler SD, Greer KL, Jaszczak RJ. Determination of mechanical and electronic shifts for pinhole SPECT using a single point source. IEEE Trans Med Imaging 2005;24:361–370. [CrossRef][Medline]
  19. Metzler SD, Bowsher JE, Greer KL, Jaszczak RJ. Analytic determination of the pinhole collimator's point-spread function and RMS resolution with penetration. IEEE Trans Med Imaging 2002;21:878–887. [CrossRef][Medline]
  20. Metzler SD, Bowsher JE, Smith MF, Jaszczak RJ. Analytic determination of pinhole collimator sensitivity with penetration. IEEE Trans Med Imaging 2001;20:730–741. [CrossRef][Medline]
  21. Chin BB, Bloomgarden DC, Xia W, et al. Right and left ventricular volume and ejection fraction by tomographic gated blood-pool scintigraphy. J Nucl Med 1997;38:942–948. [Abstract/Free Full Text]
  22. Esposito G, Santana LF, Dilly K, et al. Cellular and functional defects in a mouse model of heart failure. Am J Physiol Heart Circ Physiol 2000;279:H3101–H3112. [Abstract/Free Full Text]
  23. Tanaka N, Dalton N, Mao L, et al. Transthoracic echocardiography in models of cardiac disease in the mouse. Circulation 1996;94:1109–1117. [Abstract/Free Full Text]
  24. Wu MC, Tang HR, Gao DW, et al. ECG-gated pinhole SPECT in mice with millimeter spatial resolution. IEEE Trans Nucl Sci 2000;47:1218–1221. [CrossRef]
  25. Liang CS, Frantz RP, Suematsu M, et al. Chronic beta-adrenoceptor blockade prevents the development of beta-adrenergic subsensitivity in experimental right-sided congestive heart failure in dogs. Circulation 1991;84:254–266. [Abstract/Free Full Text]
  26. Liang CS, Fan TH, Sullebarger JT, Sakamoto S. Decreased adrenergic neuronal uptake activity in experimental right heart failure: a chamber-specific contributor to beta-adrenoceptor downregulation. J Clin Invest 1989;84:1267–1275. [Medline]
  27. Daou D, Van Kriekinge SD, Coaguila C, et al. Automatic quantification of right ventricular function with gated blood pool SPECT. J Nucl Cardiol 2004;11:293–304. [CrossRef][Medline]
  28. Vanhove C, Lahoutte T, Defrise M, Bossuyt A, Franken PR. Reproducibility of left ventricular volume and ejection fraction measurements in rat using pinhole gated SPECT. Eur J Nucl Med Mol Imaging 2005;32:211–220. [CrossRef][Medline]
  29. Tai YC, Ruangma A, Rowland D, et al. Performance evaluation of the microPET focus: a third-generation microPET scanner dedicated to animal imaging. J Nucl Med 2005;46:455–463. [Abstract/Free Full Text]
  30. Germano G, Kiat H, Kavanagh PB, et al. Automatic quantification of ejection fraction from gated myocardial perfusion SPECT. J Nucl Med 1995;36:2138–2147. [Abstract/Free Full Text]
  31. Kim SJ, Kim IJ, Kim YS, Kim YK, Shin YB, Kim DS. Automatic quantification of right ventricular volumes and right ventricular ejection fraction with gated blood pool SPECT: comparison of 8- and 16-frame gated blood pool SPECT with first-pass radionuclide angiography. J Nucl Cardiol 2005;12:553–559. [CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Movie and Figure
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chin, B. B.
Right arrow Articles by Jaszczak, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chin, B. B.
Right arrow Articles by Jaszczak, R. J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE