Left ventricular diastolic filling in patients with left ventricular dysfunction

Left ventricular diastolic filling in patients with left ventricular dysfunction

423 International Journal ojCardiology, 8 (1985) 423-436 Elsevier IJC 00273 Left ventricular diastolic filling in patients with left ventricular dy...

1MB Sizes 0 Downloads 122 Views

423

International Journal ojCardiology, 8 (1985) 423-436 Elsevier

IJC 00273

Left ventricular diastolic filling in patients with left ventricular dysfunction Steven J. Lavine, Venkataraman Krishnaswami, Morteza Amidi

David P. Shreiner

and

Section of Cardiology and Nuclear Medicine, VA Medical Center and Presbyterian-University Hospital, University of Pittsburgh, Pittsburgh, PA (Received

3 December

1984; revision accepted

29 January

1985)

Lavine SJ, Krishnaswami V, Shreiner DP, Amidi M. Left ventricular diastolic filling in patients with left ventricular dysfunction. Int J Cardiol 1985;8:423-436. The pattern of abnormal left ventricular diastolic filling and its specificity in coronary disease patients with severe left ventricular dysfunction has received little attention. We evaluated the left ventricular diastolic filling curve derived from gated blood pool scans in 21 normals, 61 coronary disease patients with ejection fractions G 30%, and 51 congestive cardiomyopathy patients with ejection fraction G 30%. The peak filling rate (PFR), peak ejection rate (PER), PFR/PER and the % stroke volume filled at l/3 of diastole (%SV-l/3 DT) and at the end of the rapid filling period (%SV-RFP) were determined for each group. The PFR and PER were reduced in both coronary disease and congestive cardiomyopathy groups. The PFR/ PER was increased in the coronary disease group (1.19 f 0.28) and congestive cardiomyopathy group (1.21 f 0.32) as compared to normals (0.93 f 0.20, P -z 0.001). A greater %SV-l/3 DT and %SV-RFP were noted in both coronary disease and congestive cardiomyopathy groups. Coronary disease and congestive cardiomyopathy patients with a mean pulmonary capillary pressure (PCP) 2 18 mm Hg had a greater PFR/PER, %SV-l/3 DT, and %SV-RFP than patients with a PCP < 18 mm Hg. An abnormal and nonspecific pattern of left ventricular diastolic filling is present in both coronary disease and congestive cardiomyopathy patients and is characterized by an increased PFR/PER, a greater %SV-l/3 DT, and a greater %SV-RFP. This pattern may be related to elevated PCPs.

(Key words: coronary artery disease: left ventricular dysfunction; filling)

Reprint requests to: Steven J. Lake, C, Pittsburgh, PA 15240, U.S.A.

0167-5273/85/$03.30

M.D., Section of Cardiology.

0 1985 Elsevier Science Publishers

VA Medical

B.V. (Biomedical

Center,

Division)

left ventricular

University

Drive

424

Introduction Resting left ventricular diastolic filling has been noted to be abnormal in patients with coronary artery disease and left ventricular dysfunction [l-4]. It is not clear whether these abnormalities in diastolic filling are specific to coronary disease or a nonspecific finding relating to left ventricular dysfunction. Patients with left ventricular dysfunction comprise a heterogeneous group of patients with ejection fractions ranging from minimal impairment to severe depression with clinically overt heart failure. The pattern of diastolic filling in coronary disease patients or cardiomyopathy patients with severe left ventricular dysfunction has received little attention. To date. Van de Werf et al. [5] and Hammermeister et al. [6], using biplane ventriculography, have noted peak diastolic filling rates as ml/set that were similar to normals in patients with cardiomyopathy, but reduced when expressed as end-diastolic volumes/set. The purpose of this study was to examine the pattern of diastolic filling in patients with coronary disease and congestive cardiomyopathy who have severe left ventricular dysfunction. Methods and Materials Patient Selection We performed gated blood pool scintigrams as part of a noninvasive evaluation of left ventricular function in 112 patients referred for cardiac catheterization for either chest pain, dyspnea, or congestive heart failure. An additional 21 patients who had no evidence of cardiac disease were also studied. Twelve-lead electrocardiograms were performed on all patients. The patients were subdivided into 3 groups. Group 1 consisted of 21 patients who had no evidence of coronary artery disease or other cardiac disease on the basis of history, physical examination, chest X-ray, echocardiography (M-mode and 2-d& mensional), and had a normal age predicted maximal exercise treadmill test. Patients taking beta- or calcium-channel blockers, or who had evidence of left ventricular hypertrophy, or a history of hypertension were excluded. The presence of left ventricular hypertrophy was assessed by voltage and other electrocardiography criteria of Romhilt and Estes [7] or by the septal or posterior wall thickness on M-mode echocardiogram > 12 mm. Group 2 consisted of 61 patients with at least one 50% lumen reducing lesion in at least one coronary artery and who had an ejection fraction < 30%. Twenty patients were referred for chest pain only, 15 for dyspnea only, and 26 for both symptoms. Previous ECG evidence of a myocardial infarction (Q waves > 40 msec) was present in 55 patients. Abnormal wall motion was noted in at least one segment in all 61 patients. Generalized hypokinesis was noted in 8 patients. Echocardiograms (M-mode and 2-dimensional) were obtained in 33 patients. Left ventricular hypertrophy was present in 6 patients as demonstrated by echocardiography or electrocardiography. Group 3 consisted of 51 patients with congestive cardiomyopathy classified either as alcohol induced or idiopathic.

425

Coronary angiography and left ventriculography were performed on 22 patients, and right-sided heart catheterization was performed on all 51 patients. Coronary arteriography was normal in all 22 patients. The remaining 29 patients had no evidence of a prior myocardial history of chest pain, no electrocardiographic infarction (Q waves < 40 msec), and no evidence of regional disease on gated blood pool scintigraphy. Echocardiograms were performed in 42 patients. Left ventricular hypertrophy was present in 2 patients by echocardiography and electrocardiography. Both patients had a history of hypertension and significant ethanol usage. Cardiac Catheterization Right-sided cardiac catheterization was performed on all 112 patients and leftheart catheterization, coronary angiography and left ventriculography in 83 patients. For each patient, the number of coronary vessels with > 50% lumen reducing lesion was determined by three trained observers. End-diastolic and end-systolic volumes of the 30” right anterior oblique ventriculogram were determined by the single plane area-length equation [8] with the Kennedy regression equation [9]. Regional wall motion was interpreted by visual assessment as normal, hypokinetic, akinetic, or dyskinetic by three independent observers. Complete agreement occurred in 74 patients. In the remaining 9 patients, a single observer determined regional wall motion. Abnormal wall motion was present when all 3 observers agreed that at least 1 segment was either hypokinetic, akinetic, or dyskinetic. Gated Blood Pool Cardiac Scintigraphy Gated blood pool scintigraphy was performed in the supine position within 48 hr of cardiac catheterization. Prior to imaging, red blood cells were labeled in vivo with 20-25 mCi of technetium-99m [lo]. Imaging was performed with a conventional gamma camera and a medium sensitivity all purpose parallel hole collimator without caudal angulation, in the anterior view, modified left anterior oblique view which best separated the left ventricle from the right ventricle, and the left lateral view. Image collection was synchronized with the QRS complex of the ECG by a computer based ECG gate (American Optical). A commercial computer system (MDS-A2) divided each cardiac cycle into 32 frames. Images were acquired in a 64 x 64 byte mode matrix for at least 250,000 counts per frame. A plot of the number of cycles collected for each RR interval was inspected. If 99% of cycles fell within rtlO% of this average RR interval, then the study was further analyzed (similar to a serial mode acquisition with a 10% window). If not, a serial mode acquisition in the “best septal” left anterior oblique projection was performed for 5 min using a f. 10% window around the mean RR interval. The data was reformatted with the aid of an ECG gate into 32 frames using forward-backward reformatting [3]. Ten patients in group 2 and 11 patients in group 3 required serial mode acquisition. Left ventricular regions of interest were derived for each frame by a semiautomatic second derivative edge detection algorithm (MUGE) applied to each of the

426

frames. A paraventricular background region of interest was selected by the algorithm. A background subtracted time activity curve was generated utilizing temporal averaging and spatial smoothing of the image sequence. The resulting curve was subjected to a 5 point smoothing routine. The ejection fraction was calculated from the background subtracted time activity curve in the standard manner. Regional wall motion was visually assessed as normal, hypokinetic, akinetic, or dyskinetic by two independent observers for each of the three views. Interobserver agreement occurred in 100% of group 1 patients. In groups 2 and 3, 11 patients produced discordant readings which were resolved by a single observer. Abnormal wall motion was present when all 3 observers agreed that at least 1 segment was either hypokinetic, akinetic, or dyskinetic. End-diastolic and end-systolic volumes were determined in each patient from the gated blood pool scintigram by a variation of a previously described count-based nongeometric method [ll-131 which was validated in our laboratory on 31 other patients. A close correlation was noted between end-diastolic radionuclide volume units and end-diastolic volume by contrast ventriculography [angiographic volume = 4.14 (radionuclide volume units) + 60.1 ml; t = 0.96, P < O.OOl].There was also a close correlation between end-systolic radionuclide volume units and end-systolic volume by contrast ventriculography [angiographic volume = 4.4 (radionuclide volume units) + 19.8 ml; r = 0.97, P < 0.011. The standard error of estimate was + 19.1 ml for end-diastole and 4 16.2 ml for end-systole. End-diastolic radionuclide volume units were calculated for each patient and converted to end-diastolic volume by use of the above regression equations. The time activity curve is analogous to a plot of volume against time. In Fig. 1, end-diastole is the highest part of the upper curve while end-systole represents the

VOLUME

CURVE

Fig. 1. The upper curve is a representation of a left ventricular volume curve derived from gated blood pool scintigraphy. The lower curve is the first derivative of the left ventricular volume curve. ED = enddiastole: ES = end-systole; PER = peak ejection rate; PFR = peak filling rate. Point A is the volume at which the peak ejection rate occurs. Point B is the volume at which the peak filling rate occurs. Point C represents the point on the derivative curve following the peak filling rate which is 20% of the peak filling rate. This point in time represents the end of the rapid filling period. The corresponding point in time (D) on the volume curve represents the volume at the end of the rapid filling period.

42-l

nadir of the curve. The time elapsed from end-systole to the following end-diastole is the diastolic time. The lower curve in Fig. 1 is the first derivative of the time activity curve and is a plot of counts/set against time. Point A on the volume curve represents the volume at which the peak ejection rate occurs. The time elapsed from end-diastole to point A is the time to the peak ejection rate. The maximum value of the first positive deflection on the derivative curve is the peak filling rate. Point B on the volume curve is the volume at which the peak filling rate occurs. The time elapsed from end-systole to point B is the time to the peak filling rate which was divided by the diastolic time to take into account the effect of heart rate. Point C on the derivative curve represents the time following the peak filling rate at which 20% of the peak filling rate occurs and its corresponding volume (point D) on the volume curve as noted by Gibson et al. [14] marks the end of the rapid filling period. The time elapsed from end-systole to point D represents the rapid filling period which was divided by the diastolic time to take into account the effect of heart rate. This period may not be a precise representation of the rapid filling period as it unavoidably contains the isovolumic relaxation period. The volume filled from end-systole to point D divided by the rapid filling period represents the mean filling rate of the rapid filling period, here defined as the mean filling rate. If after the peak filling rate, the filling rate did not fall to exactly 20% of the peak filling rate prior to atria1 filling, then the point on the derivative curve that was closest to the 20% of the peak filling rate was chosen. The filling rate fell to < 30% of the peak filling rate in all but 10 patients. In each of the 123 patients, diastasis (plateau on the filling curve following the end of the rapid filling period) was discernible and corresponded to either point D or the next point. In the other 10 patients, the volume curve was examined for diastasis which was clearly identified from the volume curve with complete agreement of 2 observers. The first point of diastasis was chosen as the end of the rapid filling period. The peak and mean filling rates were expressed as counts per second and divided by end-diastolic counts yielding the peak or mean filling rate as end-diastolic volumes per second (EDV/S). The % stroke volume filled at l/3 of diastole and at the end of the rapid filling period was obtained. Linear interpolation was used to determine the % stroke volume filled at one-third of diastole if the time of one-third of diastole fell between two points on the time activity curve. This measure was divided by the RR interval due to its high correlation (r = 0.86) with the heart rate. The ratio of the peak filling rate to the peak ejection rate (PFR/PER) was obtained as an additional filling index. This index took into account the high correlation between the peak filling rate and peak ejection rate when expressed as EDV,‘S [6]. Interobserver and intraobserver variability, for the peak filling rate and the mean filling rate were determined in a separate group of 30 patients: 9 normals, 10 with coronary disease and normal left ventricular function, and 11 with coronary disease and abnormal left ventricular function. Interobserver variability was determined by two observers independently evaluating and generating the time activity curve from the left anterior oblique view. The absolute value for interobserver variability was 0.10 + 0.11 EDV/S for the peak filling rate and 0.05 f 0.05 EDV/S for the mean

428

filling rate. Intraobserver variability was determined by two separate generations of the time activity curve by a single observer several weeks apart. The absolute value for intraobserver variability was 0.05 _+0.05 EDV/S for the peak filling rate and 0.02 k 0.03 EDV/S for the mean filling rate. Statistics All data are expressed as mean _+SD. Comparison among groups was performed by a one-way analysis of variance. If the F statistic was significant at 0.05 level, then z-tests were used to find where the difference existed. Comparison of frequency distribution between groups was performed by &i-square or Fisher’s exact cl&square test. Linear regression was performed to assess the relation between two variables. Results Clinical and Laboratory Characteristics The clinical characteristics of the coronary disease and cardiomyopathy groups are shown in Table 1. Dyspnea, edema, an S3, and left bundle branch block were more common in the cardiomyopathy group. Electrocardiographic evidence of myocardial infarction and chest pain were more common in the coronary disease

TABLE

1

Clinical

characteristics. Coronary (N=61)

disease

Cardiomyopathy (N=51)

No.

%

No.

%

s4

39 44 14 12 34 25

64 12 23 20 56 41

50 7 28 16 38 23

98 14 55 31 15 45

AWMI IWMI LBBB RBBB Cardiomegaly Cephalization l-vessel disease 2-vessel disease 3-vessel disease Left main disease

33 13 3 4 43 22 4 12 38 7

54 21 5 7 70 36 I 20 62 11

00 18 1 42 18 0 0 0 0

Dyspnea Chest pain Edema Rales s3

b b b a

;:: 35 b 2 82 35 0 0 0 0

AWMI = anterior wall myocardial infarction; IWMI = inferior wall myocardial infarction; bundle branch block; RBBB = right bundle branch block. a P < 0.05; b P -L0.001.

LBBB = left

429 TABLE

2

Age, intervals,

4% W RR (msec) DT (msec) EF (x) EDV (ml) PCP (mm Hg)

ejection

fraction,

end-diastolic

volume,

and pulmonary

capillary

pressure.

Normals (N=21)

Coronary disease (N=61)

Cardiomyopathy (N=51)

551 15 864+128 541+ 91 65& 6 151+ 30 _

60+ 11 770&163a 4825135 = 23+_ 6’ 271+ 55’ 21* 9

57+ 7 689 + 127 ce 424 k 101 cd 19k 6” 312+ 67ci 22; 9

Mean + SD. RR = average RR interval; DT = diastolic time; EF = ejection fraction; EDV = end-diastolic volume; PCP = mean pulmonary capillary pressure. a P -c 0.05, b P i 0.01, ’ P (c 0.001, normal vs. coronary disease or cardiomyopathy; d P < 0.05, e P i 0.01. ’ P < 0.001, coronary disease vs. cardiomyopathy.

group. Multivessel coronary disease was predominant, especially 3-vessel disease, in the coronary disease group. Group 1 patients had no clinical, radiographic, or electrocardiographic abnormalities. Table 2 summarizes the data concerning age, intervals, ejection fraction, end-diastolic volume, and mean pulmonary capillary pressure. A shorter RR interval, diastolic time, a larger end-diastolic volume, and a lower ejection fraction were noted in both the coronary disease and cardiomyopathy groups as compared to normals. The cardiomyopathy group had a shorter RR interval, diastolic time, a larger end-diastolic volume, and a reduced ejection fraction as compared to the coronary disease group. Systolic Ejection and Diastolic Filling Rates (Table 3) Table 3 summarizes the results for systolic ejection and diastolic filling rates. The peak ejection rate, peak filling rate, and mean filling rate, all as EDV/S, were TABLE

3

Systolic

ejection

PFR (EDV/S) PER (EDV/S) MFR (EDV/S) PFR/PER

and diastolic

filling rates.

Normals (N=21)

Coronary disease (N=61)

3.11 i 0.65 3.34io.71 1.66kO.31 0.93 * 0.20

1.83 i 0.55 1.54kO.43 0.80*0.32 1.19 + 0.28

Cardiomyopathy (N=51)

a a a a

1.69 f 0.55 1.40*0.47 0.76 + 0.34 1.21 f 0.32

’ a a a

Mean f SD. PER = peak ejection rate; PFR = peak filling rate: MFR = mean filling rate; EDV/S diastolic volumes/second, a P c 0.001 vs normals.

‘= end-

430 TABLE Diastolic

4 time intervals

TPER (msec) TPFR (msec) TPFR/DT RFP (msec) RFP/DT SSV-l/3 DT (I%) %SV-RFP (%)

and percent

filling at these intervals.

Normals (N=21)

Coronary disease (N=61)

Cardiomyopathy (N=51)

105 f48 169 i36 0.33* 0.08 297 +54 0.54* 0.11 46.1 ,13.1 73.1 _+ 9.2

101 i36 150 F51 0.325 0.12 240 &61’ 0.57* 0.14 62.3 k 23.1 ’ 80.1 + 13.1 d

98 +_50 131 +61 d 0.31+ 0.14 215 +66Cd 0.51+ 0.14 71.8 I 18.1 Cd 81.1 k14.6 b

Mean f SD. TPER = time to peak ejection rate; TPFR = time to peak filling rate; TPFR/DT = time to peak filling rate/diastolic time; RFP = rapid filling period; RFP/DT = rapid filling period/diastolic time; SW-l/3 DT = percent stroke volumes filled at l/3 of diastole (divided by the RR interval); %SV-RFP = percent stroke volume filled at the end of the rapid filling period. a P i 0.05, b P < 0.01, ’ P < 0.001,normals vs. coronary disease or cardiomyopathy; d P c 0.05,coronary disease vs. cardiomyopathy.

reduced in both coronary disease and cardiomyopathy groups. However, the PFR/PER was higher in both coronary disease and cardiomyopathy groups as compared to normals. Diastolic Time Intervals and Percent Filling at these Intervals

The time to the peak filling rate was reduced in the cardiomyopathy group, but all groups were similar when this measure was divided by diastolic time (Table 4). The rapid filling period was reduced in both coronary disease and cardiomyopathy groups, but all groups were similar when the rapid filling period was divided by the diastolic time. A greater % stroke volume filled at l/3 of diastole and at the end of the rapid filling period was noted in both the coronary disease and cardiomyopathy groups with the cardiomyopathy group having the greatest % stroke volume filled at l/3 of diastole. Diastolic Filling Pattern in Severe Left Ventricular Dysfunction

Fig. 2 demonstrates the pattern of diastolic filling in a normal patient and a patient with severe left ventricular dysfunction and an elevated mean pulmonary capillary pressure. In the patient with poor left ventricular function, the slope of the ejection and filling curves appear reduced as compared to the normal patient. However, the slope of the filling curve appears greater than the ejection curve (left side). When the time activity curves are plotted as percent stroke volume against time (right). it is evident that the rapid filling period was shorter and a greater % stroke volume was filled at the end of the rapid filling period. Also, a diastolic plateau is seen to follow the rapid filling period where little additional filling occurs. Table 5 summarizes the results of time intervals, systolic parameters, and diastolic

431

2

rl 1

1

I

TIME (MSEC)

TIME (MSEC)

Fig. 2. Left: volume curves normalized to end-diastolic volume (%EDV) derived from gated blood pool scans are shown for patient 1 with a normal ejection fraction and patient 2 with a markedly reduced ejection fraction. Right: volume curves normalized to stroke volume (&SV) are shown for patient 1 and patient 2. For patient 2, the rapid filling period (RFP,) is shorter than the rapid filling period (RFP,) for patient 1. The percentage of stroke volume filled at the end of the rapid filling period (b2) is greater in patient 2 than for patient 1 (b,).

parameters in both coronary disease and cardiomyopathy groups subdivided on the basis of the mean pulmonary capillary pressure. Groups 2A (coronary disease) and 3A (cardiomyopathy) had mean pulmonary capillary pressures < 18 mm Hg. Groups 2B and 3B had mean pulmonary capillary pressure 2 18 mm Hg. The % stroke volume filled at l/3 of diastole, the % stroke volume filled at the end of the rapid filling period, and PFR/PER were increased in both groups 2B and 3B. In groups 2A and 3A, the PFR/PER and the S stroke volume filled at the end of the

TABLE

5

Time intervals, systolic parameters. and diastolic pressure above or below 18 mm Hg. Coronary

EF (8) PCP (mm Hg) EDV (ml) RR (msec) DT (msec) TPFR/DT RFP,‘DT %SV-l/3 DT (W) SSV-RFP (%) PFR (EDV/S) MFR (EDV/S) PER (EDV,‘S) PFR/PER For abbreviations vs. 3B.

parameters

artery disease

in patients

with mean pulmonary

capillary

Cardiomyopathy

Group 2A (N=23)

Group 2B (N= 38)

Group 3A (N=14)

Group 3B (N = 37)

25 + 6 13 + 3 261 * 51 823 +114 492 +113 0.32f 0.12 0.57+ 0.13 47.4 f 18.1 74.2 + 15.1 1.72+ 0.52 0.76+ 0.26 1.58k 0.38 1.091t 0.28

22 +- 6 26 f 8’ 280 * 61 149 +132 a 471 +111 0.30* 0.11 0.56* 0.13 12.8 f 22.1= 85.1 +- 13.8 b 1.92* 0.56 0.82* 0.33 1.53* 0.49 1.25+ 0.30 a

21 +9 13 *4 284 +48 693 &89 432 +81 0.31 f 13 0.51+ 0.15 64.1 + 16.2 74.1 kll.1 1.68+ 0.63 0.81+ 0.32 1.62+ 0.50 1.04+ 0.29

18 f 6 26 & 9’ 325 F 51 a 619 rt129 420 * 79 0.34* 0.12 0.511 0.13 75.2 +_ 13.1 a 85.2 k 14.2 b 1.69* 0.45 0.74& 0.26 1.32+ 0.51 1.27 f 0.33 a

see Tables 2-4. a P

< 0.05, b P c 0.01, ’ P -c0.001; for groups 2A vs. 2B and grolips 3A

432 TABLE Incidence

6 of abnormal

diastolic

filling parameters Coronary Group

PFR/PER > 1.33 %SV-l/3 DT > 72% BSV-RFP > 91% 3 Abnormal parameters 2 Abnormal parameters only 1 Abnormal parameter only At least 1 abnormal parameter

in patients

with severe left ventricular

disease

2A

dysfunction

Cardiomyopathy Group

2B

Group

3A

Group

3B

No.

%

No.

W

No.

‘%

No.

%

3 3 2 1 1 3 5

13 13 9 4 4 13 22

17 19 13 4 11 16 31

45” 50h 34” 11 29a 42” 82c

2 2 0 0 1 2 4

14 14 0 0 I 14 21

17 20 14 4 10 19 33

46 54 38 11 27 51 89

a * h

.l c

PFR/PER = peak filling rate/peak ejection rate; %SV-l/3 DT = percent of stroke volume filled at l/3 of diastole (divided by the RR interval); %SV-RFP = percent of stroke volume filled at the end of the rapid filling period. a P c 0.05, b P i 0.01. ‘P -c0.001; for groups 2A vs. 2B and groups 3A vs. 3B.

rapid filling period were similar to normals. Also, the % stroke volume filled at l/3 of diastole was similar to normals in group 2A but remained elevated in group 3A (P < 0.01). Both coronary disease and cardiomyopathy groups were also subdivided on the basis of electrocardiographic evidence of a bundle branch block. The incidence of right bundle branch block in the coronary disease and cardiomyopathy groups and the incidence of left bundle branch block in the coronary disease group were too low for any meaningful statistical comparison. However, there was a 35% incidence of left bundle branch block in the cardiomyopathy group. No significant differences were noted between patients with and without left bundle branch block in any timing interval, systolic parameter, or diastolic parameter cited in Tables 2-4. Table 6 shows the incidence of abnormal diastolic filling parameters (PFR/PER and the % stroke volume filled at l/3 of diastole and at the end of the rapid filling period) in both coronary disease and cardiomyopathy groups. The mean + 2 SD were utilized as the upper limit of normal for each of these 3 parameters. In the normal group, the values for each of the 3 parameters were less than the upper limit of normal for each patient. A low incidence of abnormalities (not significant) for any 1 parameter was noted in both groups 2A and 3A. However, at least one abnormality was noted in 22% (P -C0.05 as compared to normals) of coronary disease and 21% (P < 0.05) of cardiomyopathy patients. A significantly higher incidence of abnormalities was noted in groups 2B and 3B as compared to 2A and 3A. An incidence of at least 34% was seen in groups 2B and 3B for any 1 parameter. At least 1 abnormal parameter was noted in 82% of group 2B and 89% of group 3B patients. Relation between RR Interval, End-Diastolic Pressure, Ejection Fraction, and Diastolic Filling Parameters (Table 7)

In normals, the RR interval was unrelated to any diastolic filling parameter. The ejection fraction was well correlated with the peak filling rate.

433

TABLE

7

Correlation parameters.

between

RR interval.

Normals

PFR (EDV/S) PFR/PER %SV-l/3 DT BSV-RFP

pulmonary

( I%’= 21)

capillary

pressure.

ejection

fraction.

and diastolic

filling

CM(N=51)

CD(N=61)

RR

EF

RR

PCP

EF

- 0.24 0.27 0.01 0.04

0.16 h -0.33 - 0.24 - 0.23

-0.36 ’ 0.01 0.07 - 0.20

0.18 0.47 h 0.36 ’ 0.37 d

- 0.05 -0.23 - 0.22

0.63 ~’

RR

PCP

EF

-0.51 h - 0.08 0.18 - 0.26

0.01 0.57 h 0.39 .J 0.50 h

0.67 h - 0.04 - 0.01 0.07

CD = coronary disease; CM = cardiomyopathy; RR = RR interval; PCP = mean pulmonary capillary PFR = peak filling rate: EDV/S = end-diastolic volumes/set: pressure: EF = ejection fraction; PFR/PER = peak filling rate/peak ejection rate: %SV-l/3 DT = percent stroke volume filled at l/3 of diastole; %SV-RFP = percent stroke volume filled at the end of the rapid filling period. a P < 0.01: hP
In coronary disease and cardiomyopathy patients. the peak filling rate was modestly to moderately related to the RR interval and well related to the ejection fraction. The mean pulmonary capillary pressure was modestly to moderately correlated with the PFR/PER and the o/cstroke volume filled at l/3 of diastole and at the end of the rapid filling period.

Discussion We evaluated the pattern of diastolic filling in coronary disease and cardiomyopathy patients with severe left ventricular dysfunction (ejection fractions G 30%). The peak filling rate, mean filling rate, and peak ejection rate as EDV/S in this study were reduced to a similar degree in both coronary disease and cardiomyopathy groups as compared to normals. If the peak filling rates were expressed as ml/set. no significant differences were noted between normals (476 f 109 ml). coronary disease patients (455 + 150 ml) and cardiomyopathy patients (471 + 172 ml). This finding has also been previously reported in cardiomyopathy patients [5,6]. However, an abnormal pattern of diastolic filling was noted in both coronary disease and cardiomyopathy groups (Fig. 2) and was characterized by an elevated PFR/PER, an increased % stroke volume filled at l/3 of diastole, and an increased % stroke volume filled at the end of the rapid filling period. A diastolic plateau was also noted with little additional filling after the rapid filling period. An example of this pattern is seen in Fig. 2. This abnormal pattern of diastolic filling was found in both coronary disease and cardiomyopathy subgroups with mean pulmonary capillary pressure > 18 mm Hg. Table 6 shows the incidence of these abnormalities using the mean + 2 SD of normals as the upper limit of normal. At least one abnormality was seen in 82 and 89% of coronary disease and cardiomyopathy patients with a mean pulmonary capillary pressure > 18 mm Hg, 22 and 21% of coronary disease and cardiomyopathy patients with a mean pulmonary capillary pressure < 18 mm Hg. and 0% of normals. A significant relationship was noted between the mean pulmonary

434

capillary pressure and the PFR/PER, % stroke volume filled at l/3 of diastole, and the % stroke volume filled at the end of the rapid filling period. Of importance, each of these 3 parameters was unrelated to heart rate or ejection fraction.

Previous Literature Observations To date, this pattern of diastolic filling has not been fully recognized. Van de Werf et al. [5] and Hammermeister et al. [6] have noted similar peak filling rates as ml/set in cardiomyopathy patients as compared to normals. Carrol [15] noted similar peak and mean filling rates as ml/set as compared to normals during the first l/2 of diastole in coronary disease patients who developed exercise induced ischemia. However, most of diastolic filling occurred during the rapid filling period with little additional filling following the rapid filling period. This pattern of diastolic filling was associated with a marked increase in end-diastolic pressure, a slight increase in end-diastolic volume (still in the normal range), and a reduction in the ejection fraction to a borderline normal level [15].

Theoretical Explanations for this Abnormal Pattern of Diastolic Filling Our data provide no definite explanation for this abnormal pattern of diastolic filling. A potential explanation may be found in the significant correlation between mean pulmonary capillary pressure and the PFR/PER, % stroke volume filled at l/3 of diastole, and the % stroke volume filled at the end of the rapid filling period. Although these correlations are only moderate, most of the abnormalities in the parameters describing this altered pattern of diastolic filling were found in the coronary disease and cardiomyopathy subgroups with a mean pulmonary capillary pressure > 18 mm Hg. Yellin et al. [16] have noted that early left ventricular filling may be related to the magnitude of atrial-ventricular pressure difference and the rate at which these differences develop at the time of the mitral valve opening. It can be anticipated that the left atria1 pressure at the time of mitral valve opening will be elevated if the mean pulmonary capillary pressure is also elevated. The extent and rate of left ventricular pressure decline in early diastole may also be an important factor in determining the rate of early diastolic filling [17,18]. The increased PFR/PER could be explained if the elevation of the left atria1 pressure overshadowed the effect of a prolonged time constant of left ventricular pressure decline [19,20] in patients with left ventricular dysfunction. In the case of coronary disease patients, an elevated left atria1 pressure may also overcome the tendency of asynchronous diastolic filling to reduce the rate of early diastolic filling [21,22]. More complete diastolic filling at the end of the rapid filling period can be potentially explained by the fact that the instantaneous compliance of the left ventricle decreases as the volume and pressure of the left ventricle increase resulting in greater impedance to left ventricular filling [23]. This may be of considerable importance in patients with large left ventricles with high filling pressures. Since simultaneous left ventricular filling pressures and diastolic filling rates were not obtained, the effect of an elevated left atria1 pressure and mid-diastolic pressure

435

volume relations on diastolic filling cannot be ascertained. The above explanations for this abnormal pattern of diastolic filling are at best speculative. Furthermore, simultaneous gated blood pool and hemodynamic studies during an intervention that affected the mean pulmonary capillary pressure would be necessary to assess the role that left ventricular filling pressure plays in the production of this abnormal pattern of diastolic filling. Specificity

of this Pattern of Diastolic

Filling

The PFR/PER and % stroke volume filled at the rapid filling period were found to be equally and cardiomyopathy subgroups. This abnormal specific for coronary disease and is probably a ventricular dysfunction and possibly elevated left

l/3 of diastole and at the abnormal in the coronary pattern of diastolic filling nonspecific finding related ventricular filling pressures.

end of disease is not to left

Clinical Implications An abnormal pattern of left ventricular diastolic filling was noted in coronary disease and congestive cardiomyopathy patients. This pattern of diastolic filling was much more prevalent in the subgroup of patients with elevated mean pulmonary capillary pressure (a 18 mm Hg) as opposed to the subgroup of patients with mean pulmonary capillary pressures < 18 mm Hg. The ability to detect a change in the pattern of diastolic filling in the therapy of congestive heart failure or with any intervention designed to lower the left ventricular filling pressure would be advantageous. Our data do not allow us to suggest an alteration in the pattern of diastolic filling with lowering of the left ventricular filling pressure. Clearly, further work is needed to determine the role that the left ventricular filling pressure plays in the production of this abnormal pattern of diastolic filling and whether this pattern can be seen in other valvular or myocardial diseases.

Acknowledgement We would like to thank manuscript.

Mira Betts for her assistance

in the preparation

of this

References 1 Reduto LA, Wickemeyer WJ. Young JB, et al. Left ventricular diastolic performance at rest and during exercise in patients with coronary artery disease. Circulation 1981;63:1228-1237. 2 Polak JF, Kemper AJ, Bianco JA, Parisi AF. Tow DE. Resting early diastolic filling rate. A sensitive index of myocardial dysfunction in patients with coronary artery disease. J Nucl Med 1982;23:471-478. 3 Mancini JGB, Slutsky RA. Norris SL, Bhargava V. Ashbum WL, Higgins CB. Radionuclide analysis of the peak filling rate, filling fraction. and time to peak filling rate. Am J Cardiol 1983;51:43-51. 4 Bonow RO. Bacharach SL, Green MV, et al. Impaired left ventricular diastolic filling in patients with coronary artery disease: assessment with radionuclide angiography. Circulation 1981:64:315-323. 5 Van de Werf F, Bael A, Geboers J, Minten J, Willems J, DeGeest H. Kesteloot. H. Diastolic properties of the left ventricle in normal adults and in patients with third heart sounds. Circulation 1984:69:1070-1078.

436 6 Hammermeister KE, Warbasse JR. The rate of change of left ventricular volume in man. II. Diastolic events in health and disease. Circulation 1974;494:739-747. 7 Romhilt DW, Estes EH. A point score system for ECG diagnosis of left ventricular hypertrophy. Am Heart J 1968;75:752-759. 8 Sandler H, Dodge HT. The use of single plane angiocardiograms for calculation of left ventricular volume in man. Am I!cart J 1968;75:325-334. 9 Kennedy JW, Trenholine SE. Kasser JS. Left ventricular volume and mass from single plane cineangiocardiograms: a comparison of anteroposterior and right anterior oblique methods. Am Heart J 1970;80:343-352. 10 Thrall JH, Freitas JE, Swanson D, Rogers WL, Clare JM, Brown ML. PHB. Clinical comparison of cardiac blood pool visualization with technetium 99m human serum albumin. J Nucl Med 1978;19:796-803. 11 Slutsky R, Karliner J, Ricci D, et al. Left ventricular volumes by gated equilibrium radionuclide angiography: a new method. Circulation 1979;60:556-564. 12 Dehmer CJ, Lewis SE. Hillis LD, et al. Nongeometric determinants of left ventricular volume from equilibrium blood pool scans. Am J Cardiol 1980;45:293-300. 13 Massie BM, Kramer BL. Gertz EW. Henderson SG. Radionuclide measurement of left ventricular volume: comparisons of geometric and count-based methods. Circulation 1982;65:725-730. 14 Gibson DG, Brown D. Measurement of instantaneous left ventricular dimensions and filling rate using echocardiography. Br Heart J 1973;35:1141-1149. 15 Carrol JD. Hess OM. Hirzl HO, Krayenbuehl HP. Dynamics of left ventricular filling at rest and during exercise. Circulation 1983;68:59-67. 16 Yellin EL, Sonnenblick EH, Frater RWM. Dynamic determinants of left ventricular filling: an overview. In: Baan J, Arntzenius AC, Yellin EL. eds. Cardiac dynamics. The Hague: Martinus Nijhoff, 1980;145-150. 17 Fioretti P, Brower RW, Miesler GT, Serroys PW. Interaction of left ventricular relaxation and filling during early diastole in human subjects, Am J Cardiol 1980;46:197-203. 18 Magorien DJ, Shaffer P, Bush C, et al. Hemodynamic correlates of timing intervals, ejection rate, and filling rate derived from radionuclide angiographic volume curve. Am J Cardiol 1984;53:567-571. 19 Hiroto Y. A clinical study of left ventricular relaxation. Circulation 1980;62:756-763. 20 Rousseau MF. Veriter C. Deny JMR. Brausseur L. Poleur H. Impaired early left ventricular relaxation in coronary artery disease: effects of intracoronary nifedipine. Circulation 1980;62:764-772. 21 Miller TR, Goldman KJ, Sampathkumaran KS, Biello DR, Ludbrook PA, Sobel BE. Analysis of cardiac diastolic function: application in coronary artery disease. J Nucl Med 1983;24:2-7. 22 Yamagishi T, Ozaki M. Kumada T. et al. Asynchronous left ventricular diastolic filling in patients with isolated disease of the left anterior descending coronary artery: Assessment with radionuclide ventriculography. Circulation 1984;69:933-942. 23 Gaasch WH, Levine HJ, Quinones MA, Alexander J. Left ventricular compliance: mechanism and clinical implications. Am J Cardiol 1976:38:645-653.