|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Medicine, Cardiovascular Division (Drs. Aepfelbacher, Yeon, Ho, and Danias), and Radiology, Nuclear Medicine Division (Dr. Parker), Beth Israel Deaconess Medical Center, and Harvard Medical School, Boston, MA.
Correspondence to: Peter G. Danias, MD, PhD, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215; e-mail: pdanias{at}caregroup.harvard.edu
| Abstract |
|---|
|
|
|---|
Design: Retrospective analysis.
Setting: A major university teaching hospital.
Patients: A consecutive series of 194 patients with good quality stress SPECT and 2D-ECHO studies performed within 1 day of each other and no significant interim cardiac events.
Measurements and results: RV size and function were visually assessed by SPECT/ECG-gated SPECT and 2D-ECHO. RV wall thickness was visually assessed by SPECT and measured in mm in end-diastole by 2D-ECHO. Of 142 patients with normal RV cavity size by SPECT, 134 patients (94%) had normal RV cavity size by 2D-ECHO. However, of 52 patients with RV dilation by SPECT, only 9 patients (17%) had RV dilation by 2D-ECHO. A perfusion abnormality in the right coronary artery territory was significantly associated with RV dilation by SPECT (p < 0.005) and 2D-ECHO (p < 0.05). Among 150 patients with ECG-gated SPECT, only 2 patients had abnormal RV systolic function, as compared with 18 patients by 2D-ECHO. RV wall thickness measurements by SPECT and 2D-ECHO did not correlate.
Conclusions: For normal interpretations regarding RV cavity size, wall thickness, and systolic function, there is good agreement between 99mTc stress SPECT/ECG-gated SPECT and 2D-ECHO. However, there is poor overall agreement between gated SPECT/ECG-gated SPECT and 2D-ECHO regarding the presence of RV dilation, hypertrophy, and systolic dysfunction.
Key Words: ECG-gated single-photon emission CT echocardiography right ventricle single-photon emission CT
| Introduction |
|---|
|
|
|---|
Single-photon emission CT (SPECT) and ECG-gated SPECT with 99mTc have been extensively validated for simultaneous assessment of left ventricular (LV) perfusion and function,11 12 13 14 and have been shown to correlate well with other imaging modalities for evaluation of both global and regional LV systolic function.15 16 17 Since visualization of the RV with 99mTc based agents is possible for the majority of patients, ECG-gated 99mTc SPECT imaging may have a role for the assessment of RV perfusion,18 19 size, and function.20 21 The utility of 99mTc SPECT/ECG-gated SPECT for evaluation of the RV has not been reported in a large cohort of patients. In the present study, we assessed the RV cavity size, wall thickness, and systolic function by stress 99mTc sestamibi SPECT/ECG-gated SPECT in a group of consecutive patients with known or suspected coronary artery disease, and explored the correlation of scintigraphic data with measures of RV structure and function by 2D-ECHO.
| Materials and Methods |
|---|
|
|
|---|
|
SPECT Interpretation
All stress SPECT images were interpreted by consensus of two experienced readers who had no knowledge of clinical or 2D-ECHO data. RV cavity size was assessed semiquantitatively on a 0 to 3 scale (0 = normal size, 1 = mildly dilated, 2 = moderately dilated, 3 = severely dilated). RV wall thickness was evaluated visually from the apparent wall thickness and intensity of the RV activity on a 0 to 3 scale (0 = normal, 1 = mildly increased, 2 = moderately increased, 3 = severely increased). Adequate ECG-gated SPECT images were used to grade RV systolic function on a 0 to 3 scale (0 = normal function, 1 = mild dysfunction, 2 = moderate dysfunction, 3 = severe dysfunction). For patients with ECG-gated SPECT, RV wall thickness was also assessed using the end-diastolic frame and scored visually on the previously described 0 to 3 scale. The stress perfusion of the distribution of the right coronary artery (inferior, inferoseptal, and inferoapical LV walls) was also visually assessed and characterized as "normal" or "abnormal" (moderate reduction of counts or greater).
2D-ECHO
Resting 2D-ECHO was performed in the left lateral decubitus and supine positions. The studies were performed by experienced sonographers and routinely included M-mode (parasternal view) and two-dimensional views from the parasternal, apical, suprasternal, and subcostal windows. Continuous, pulsed wave, and color Doppler were performed routinely to assess valvular competence. When tricuspid regurgitation was detected, the peak tricuspid regurgitant velocity (TRV), a correlate of RV systolic pressure, was quantified using continuous-wave Doppler recordings from the parasternal and apical windows.
2D-ECHO Interpretation
All 2D-ECHO images were interpreted by consensus of two experienced readers who had no knowledge of clinical or SPECT data. RV cavity size was assessed visually on a 0 to 3 scale (0 = normal size, 1 = mildly dilated, 2 = moderately dilated, 3 = severely dilated). RV wall thickness was measured in millimeters in end-diastole from the parasternal long-axis image. RV hypertrophy was defined as RV wall thickness > 5 mm.8
RV systolic function was assessed semiquantitatively on a 0 to 3 scale (0 = normal function, 1 = mildly reduced, 2 = moderately reduced, 3 = severely reduced function).
Statistical Analysis
Statistical analysis was performed on a personal computer using SAS for Windows (version 6.12; SAS Institute; Cary, NC). Data for continuous variables are expressed as mean ± 1 SD. Agreement between 2D-ECHO and SPECT for RV cavity size was evaluated using the McNemar test. The correlation between RV wall thickness and TRV was assessed using linear regression analysis. We also examined the relationships between TRV and RV wall thickness or cavity dilation using analysis of variance. Finally, we performed a subgroup analysis in patients who had stress perfusion abnormalities in the right coronary artery territory, and used
2 tests to evaluate the association of abnormal stress perfusion with RV size and function. All tests were two sided, and p < 0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
|
2D-ECHO
By 2D-ECHO, 12 of 194 patients (6%) had mild RV dilation and 5 patients (3%) had moderate or severe RV dilation (Table 2)
. Four patients (2%) had RV hypertrophy. Mild RV systolic dysfunction was present in 9 patients (5%), and moderate or severe RV dysfunction was seen in 14 patients (7%). TRV was quantifiable in 127 patients (65%) and ranged from 1.6 to 4.5 m/s.
Association and Agreement Between 99mTc SPECT/ECG-Gated SPECT and 2D-ECHO
The vast majority of cases in which SPECT characterized the RV cavity size as normal also had normal RV cavity size by 2D-ECHO (134 of 142 patients, 94%). None of the patients with normal RV cavity size by SPECT had moderate or severe RV dilation by 2D-ECHO; however, overall agreement between the two imaging modalities in classifying RV cavity size was poor (k value = 0.13), as 43 of the 52 patients with RV dilation by SPECT had normal RV cavity size by 2D-ECHO. Conversely, 8 of the 17 patients with RV dilation by 2D-ECHO had normal RV cavity size by SPECT (Table 2)
.
There was also poor agreement between ECG-gated SPECT and 2D-ECHO regarding RV systolic function: only one of the two patients with abnormal RV systolic function by ECG-gated SPECT had abnormal RV function by 2D-ECHO. Conversely, of the 18 patients with systolic RV dysfunction by 2D-ECHO and adequate ECG-gated SPECT, only 1 patient had RV systolic dysfunction by ECG-gated SPECT.
No significant correlation was found between RV wall thickness by 2D-ECHO and by SPECT; however, a significant association was found between TRV and RV wall thickness by SPECT, as determined either from the SPECT images (127 patients, p < 0.0005) or the end-diastolic frames of the gated SPECT images (95 patients, p < 0.0005) [Fig 1 ]. TRV also correlated significantly but weakly with RV wall thickness by 2D-ECHO (p < 0.0005, r2 = 0.10) and RV cavity size by 2D-ECHO (p < 0.05) but not with RV cavity size by SPECT (p = 0.1) [Fig 2 ].
|
|
| Discussion |
|---|
|
|
|---|
In our study, normal findings regarding RV cavity size and systolic function by SPECT/ECG-gated SPECT agreed well with 2D-ECHO findings; however, there was poor agreement between SPECT/ECG-gated SPECT and 2D-ECHO regarding the presence and severity of RV dilation and dysfunction. With SPECT, the RV cavity size was interpreted as dilated more than three times as often as with 2D-ECHO, and the majority of even moderately or severely enlarged RVs by SPECT were interpreted as normal by 2D-ECHO. Because neither of these imaging modalities can be considered a "gold standard," it is unclear whether SPECT overestimates RV cavity size, 2D-ECHO underestimates it, or both. The fact that RV systolic pressure (as assessed by TRV) correlated with severity of RV dilation by 2D-ECHO but not by SPECT suggests that SPECT likely overestimated RV cavity size in some patients. The discrepancies between the two imaging modalities may in part be due to differences in visualization of the anatomically complex RV. With SPECT, the basal RV free wall is well visualized (short-axis and horizontal long-axis views), while the apical region is frequently not distinguishable from the adjacent intraventricular septum due to poor resolution and partial volume averaging. This may lead to overestimation of RV cavity size, because of the relatively large appearance of the RV in the basal segments, compared with the LV. With 2D-ECHO, RV cavity size is assessed from the parasternal, apical four-chamber, and subcostal views, which provide variable visualization of the RV outflow, and basal and apical RV. Even minor changes in rotation and angulation of the ultrasound transducer can introduce significant variability in apparent RV cavity size; therefore, assessment of RV cavity size by 2D-ECHO and SPECT are both subject to errors, and may be based on different parts of the RV, producing apparently discordant results in some patients.
Although RV wall thickness by SPECT was not significantly associated with RV wall thickness by 2D-ECHO, radioisotopic RV wall thickening was significantly associated with TRV, a functional measure of RV pressure overload. We arrived at the same conclusion when we used only the diastolic frames of the ECG-gated SPECT images, indicating that systolic motion and thickening of the RV walls are likely not contributing significantly to this association. Our findings are in line with previous studies reporting a good correlation between RV tracer uptake with pulmonary arterial pressures in patients with congenital heart disease26 27 28 or pulmonary diseases.25 29 30
Compared to 2D-ECHO, RV dysfunction was rarely identified by ECG-gated SPECT, suggesting that visual interpretation of ECG-gated SPECT may not be sensitive enough to detect small changes of RV systolic function. Since 2D-ECHO and SPECT tend to visualize different parts of the RV, regional differences in activation pattern, contractility, and contribution to stroke volume7 may account in part for the discordance in the assessment of RV systolic function. Furthermore, SPECT image acquisition occurs over multiple cardiac cycles and disregards both cardiac and respiratory motion. While the blurring effect caused by averaging has not been a significant limitation for assessment of LV function, assessment of RV function may be affected more, due to the thinner wall and more complex shape and contraction pattern.
The clinical implications of our study are that in patients who are evaluated with 99mTc stress SPECT/ECG-gated SPECT for known or suspected coronary artery disease, a normal interpretation of RV size and function is a reliable finding. When additional specific workup for RV pathology is indicated for patients with normal-appearing RV on SPECT/ECG-gated SPECT, it is unlikely that a 2D-ECHO will be of added value, and one should reserve to a more definitive test, such as cardiac MRI. Similarly, for abnormal SPECT/ECG-gated SPECT findings of RV size and function, a definitive test would likely also be indicated, since a discrepant 2D-ECHO result would not rule out RV pathology. Abnormalities of perfusion of the right coronary artery territory should alert toward possible RV abnormalities, as these appear to be associated.
Limitations
Our study was performed without a "goldstandard" for evaluation of the RV; however, as in clinical practice 2D-ECHO is frequently performed to assess RV size and function, the question we tried to answer is whether ECG-gated SPECT can provide the same information as 2D-ECHO in patients who are referred for stress testing. For the same reason, we elected to evaluate a consecutive series of patients who were not preselected for disorders affecting the RV. While this accounted for a relatively low number of abnormal RV findings, our sample is representative of patients being evaluated by stress SPECT for known or suspected coronary artery disease, a population for whom the clinical value of ECG-gated SPECT for assessment of the RV is sought.
In our study, the two imaging modalities were not performed simultaneously. While 2D-ECHO images were obtained at rest, SPECT/ECG-gated SPECT images were obtained 20 to 60 min after stress (as is routinely done in clinical practice), which may affect RV size and function due to altered loading conditions, changes in autonomic tone, or induction of myocardial ischemia. We tried to minimize the potential of a significant interim change of RV size and function by including only patients who had both studies performed within 1 day, and by excluding patients with significant clinical events. However, the precise matching of conditions for the two tests is impractical.
| Conclusions |
|---|
|
|
|---|
| Footnotes |
|---|
This study was supported in part by a research grant from Dupont Radiopharmaceuticals (CG #99031).
Received for publication September 30, 2002. Accepted for publication November 11, 2002.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. R. Prince, P. O. Alderson, and H. D. Sostman Chronic Pulmonary Embolism: Combining MR Angiography with Functional Assessment Radiology, August 1, 2004; 232(2): 325 - 326. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |