M.D., MASAMI NAKAWUKA M.D., JUNYA MAEDA, M.D., MASAKO MATSUDA, M.D. ..... Yamagishi T, Ozaki M, Kumada T, Ikezono T, Shimizu T, Ogawa.
Clin. Cardiol. 16,816-822 (1993)
Assessment of Right Ventricular Diastolic Filling in Patients with Coronary Artery Disease TAKASHI YAMAGISHI, M.D., YASUO MNSUDA, M.D., MASAMI NAKAWUKA M.D., JUNYAMAEDA,M.D., MASAKOMATSUDA,M.D. Tokuyama Central Hospital and Yamaguchi University School of Medicine, Yamaguchi, Japan
Summary: To assess right ventricular (RV) diastolicfilling in coronary artery disease (CAD), with special reference to the involved lesions of the coronary arteries and left ventricular (LV) systolic function, gated radionuclide ventriculography was performed at rest in 106 patients with single-vesselCAD. Based on the site of coronary arterial involvement, patients were classified into three groups: left anterior descending CAD, right CAD, and left circumflex CAD. Patients in each group were further subdivided according to normal or decreased LV ejection fraction, resulting in six groups. Seventeen normal subjects were examined as a control group. Tiie-activity and its fmtderivative curves were computed for the right and left ventricles. RV systolic function was normally preserved in all six groups, even when LV systolic function was damaged severely. The ratio of peak RV filling rate to peak RV ejection rate was significantly decreased in all six groups compared with that in control subjects, indicating that RV filling was impaired in patients with CAD. The ratio was below the lower limit of normal in 14 (23%) of 62 patients with normal LV systolic function and in 13(30%) of 44 patients with impaired LV systolic function. None of the control subjects showed a decreasedratio of peak RV filling rate to peak RV ejection rate. Thus, in patients with CAD, RV filling is impamxi, which may be independent of the site of coronary arterialinvolvementand of the LV or RV systolic function.
Introduction Due to the fact that the cardiovascular system is a closed loop with the two ventricles in direct apposition, the structure and function of the right and left ventricles are interrelated in a complex manner through the myocardium, the circulatoryconnections between the two ventricles,and the coronary arterialperfusion. Some myocardial fibers span the ventricular chambers and so contributeto the function of both ventricles. Regarding arterial perfusion, one coronary artery usually perfuses both ventricles. Thus, it is reasonable to expect that an abnormal morphology andor function of one ventricle would adversely influence the function of the other. Many investigators have observed abnormalities in diastolic left ventricular (Lv) filling at rest in patients with coronary artery disease (CAD);'" however, little is known about the pvalence and extent of the abnormalitiesof diastolic right ventricular (RV) filling in such patients at rest." Gated equilibrium radionuclide ventriculography now provides a simple and noninvasive method of measuring systolic emptying and diastolic filling simultaneously in both ventricles. Using radionuclide ventriculography,our purpose was to investigatewhether the systolic emptying and diastolic filling of the right ventricle at rest would be influenced by LV dysfunction in patients with CAD. Particular emphasis was placed on the assessment of diastolic RV filling with special reference to obstructive lesions of the coronary arteries.
Key words: right ventricular diastolic fdling, coronary artery disease, gated radionuclide ventriculography
Methods Patient Selection
Address for reprints: Yasuo Matsuda, M.D. Okataya, Konan Ube, Yarnaguchi 759-02 Japan Received: March 2, 1993 Accepted: May 28, 1993
This was a retrospective study. The study group consisted of 106consecutivepatients (70 men and 36 women, aged 36-74 years) with CAD. In addition, 17 normal subjects, 9 men and 8 women aged 36-67 years, served as controls.The diagnosis of CAD was made from the medical history and from results of electrocardiography, echocardiography, coronary angiography, left ventriculography,and thallium-201 scintigraphy. Of the 106patients, 66 had a previous myocardial infarction (ryn), which was anterior in 32, inferior in 17, and posterolateralin
T. Yamagishi et al.: Right ventricular filling in C A D
17. The possibility of RV infarction is much higher with inferior infarction and this could affect both systolic and diastolic RV indices. In all patients with suspected RV infarction, a Swan-Ganz catheter was introduced during the course of the acute infarction. None had documented evidence of RV infarction. Radionuclide ventriculography was performed at least 1 month after the onset of MI when the patients were clinically stable. The remaining 40patients suffered from angina pectoris. All patients had significant organic stenosis (275% luminal diameter) in the proximal or midportion of only one of the threemajor coronary arteries. Patients were classified into threep u p s based on the site of coronary arterial involvement: left anterior descending coronary artery (LAD) disease, right coronary artery (RCA) disease, and left circumflex coronary artery (LCx) disease. The radionuclide-derived LV ejection fmction of the 17 conaol subjectswho participated in this study was 58 f 6% (range 75-50%). The lower limit of normal for LV ejection fraction was defined as the mean minus 2 standard deviations (SDs) (58 - 2 X 6 =a%)? This choice for the 4.6% cutoff introduced from the 17 control subjects is similar to that of other investigators.' Patients in each p u p were further subdivided according to their LV ejection fraction, measured by radionuclide ventriculography, as 246% and c 46%, resulting in six p u p s . The clinical characteristicsof each p u p are presented in Table 1. Cardiac medicationswere discontinuedat least 48 h before radionuclide ventriculography except for sublingual nitroglycerin, which was allowed for a n p a l attacks. All patients were in normal sinus rhythm and had no conduction disturbances. All of the 17 control subjects had atypical chest pain with normal coronary arteriograms and normal left ventriculograms. None showed clinical or hemodynamic evidence of cardiopulmonary disease as assessed by history, physical examination, exercise thallium-201 scintigraphy, electrocardiography, and chest radiography. The control group did not differ significantly from the six p u p s of patients with regad to mean age, heart rate, or arterial blood pressure. Excluded from study were individualswith congenitalheart disease, hypertensive heart disease, arrhythmia, valvular heart disease, or cardiomyopathy.
817
Radionuclide ventriculography was performed within 7 days of cardiac catheterization. Radionuclide Ventriculography Radionuclideventriculographywas performed using a conventional gamma camera (GCA601E, Toshiba) equipped with a high resolution, all-purpose, parallel-hole collimator. Subjects were studied while they were resting quietly in supine position. They received 20-25 mCi (740-925 MBq) of %-technetiumlabeled human serum albumin and, after the radionuclide had equilibrated with the intravascular space (about 10 min), the camera position was adjusted in the modified left anterior oblique projection. Obliquity varied between 40and 60" depending on the greatest separation between right and left ventricles. In addition, a 15" caudal tilt of the detector was used to improve the separation between left atrium and left ventricle. A total of 12,000to 15,000k counts was then obtained for each subject in a list-mode data acquisition format on a magnetic disc using a digital computer. We recorded those photoevents falling within a 20%window and centered on the photopeak of technetium-%. A histogram of the acquired cardiac cycle lengths was constructed and any irregular beats were excluded. The average cardiac cycle length was then determinedand separated into 30 equal frames. The original list-mode data were reformated into 30 equal frames with a 64 X 64 matrix using a combined forward and reverse gating method from the R wave. Both right atrium and pulmonary outflow tract usually overlap the right ventricle; these are difficult to exclude completely when the RV region of interest is assigned. To minimize such overlap, the global radionuclide ventriculographic data were fitted onto a pixel-by-pixel basis with the first harmonic approximation and the Fourier phase and amplitude functional images were constructed. These functional images for both right and left ventricles were used to help outline manually the RV and LV regions of interest (Fig. 1). Both RV and LV regions of interest were then automatically assigned frame-by-frame, using the variable regions of interest method at a 60% isocount contour
TABLE I Clinical characteristicsof 17 control subjects and 106 patients with coronary artery disease No. Normal LAD-A
-B RCA-A
-B LCX-A -B
Mean
Heart
age
rate
Blood pressure systoliddiastolic ( M g )
of patients
Sex (M/F)
(Yew)
(beats/min)
17 29 26 16 10 17 8
918 20/9 17/9 1115
57k 10 60* 1 1 61 f 11 61 *7 62k 12 61 * 8 60* 1 1
65 13 62* 10 69*9 69* 13 66* 10
713 1116 44
*
60*8 76k 12
118*12/78*9 120 f 16/77f 9 120*24/77* 12 121 *20/78k7 127k21/84k15 116 f 14/76 f 14 127 f 2W82 f 4
Abbreviations: M = male, F = female,LAD = left anteriordescending comnary artery disease, RCA =rightcoronary artery disease,LCx = left circumflex coronary artery disease, A = left ventricular ejection fraction 246%. B = left ventricular ejection fraction c46%, 46%= lower limit of the normal value of the left ventricular ejection fraction, defined as the mean minus 2 SDs derived from 17 control subjects.
818
Clin. Cardiol. Vol. 16. November 1993 CAD
Controls
FIG.1 Fourier phase and amplitude functional images for both the right (RV) and left ventricles (LV) were used to help outline manually the RV and LV regions of interest. 40t
within the previously mentioned (manuallyoutlined) region of interest. The background regions of interest were also automatically assigned outside the lower lateral portions of the left and right ventricles. For both ventricles we generated background-correctedtime activity and their first derivative curves (dV/dt).A series of systolic and diastolic phase indices was obtained for both ventricles as described below (Fig. 2). The systolic phase indices consisted of (1) ejection fraction (EF, %), (2) peak ejection rates normalized to end-diastolic counts and smkecounts (PER, end-diastolic countds and stroke count&), (3) time to end-systole (TES, ms): time interval between the electrocardiographic R wave and the fi-ame with minimum counts at which dV/dt value is zero. The diastolicphase indices consisted of (1) peak filling rates normalized to end-diastoliccounts and stroke counts (PFR, enddiastolic countds and stroke counts/s;(2) peak filling rawpeak ejection rate (PFRFER): ratio used to normalize the peak filling rate for the peak ejection rate and to cancel the influence of end-diastolic counts on either peak filling or peak ejection (3) time to peak filling rate (TPFR,ms): time elapsed from end-systoleto the peak filling rate. The reproducibility of these variables in measuring LV function has been reported and validated in patients with and without regional wall motion abnormalities?, The reproducibility in measuring RV systolic or diastolic function was determined in I2 patients with various heart diseases. l h o separate radionuclide studies were performed 15 min apart on resting patients following a single injection of technetium-%. An excellent correlation was observed between the RV systolic and diastolic variables obtained from the first and second studies (r 20.92).
T
t
40
T
FIG.2 Representativetime-activitycurves and their first derivative curves in the right (RV) and left ventricles (LV) taken from a control subject and a patient (CAD) with isolated disease of the left anterior descending coronary artery. Early diastolic filling was impaired in both ventricles in the patient with CAD (arrows). PER = peak ejection rate, TES = time to end-systole,PFR = peak filling rate, TPFR = time to peak filling rate.
analysis, Student’s ?-test for unpaired data, and one-way analysis of variance with multiple comparisons, as appropriate.The level of statistical significance was p
