View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

JACC Vol. 13. No. S 1007 April 1989:3007-Id

CLINICAL STUDIES

Differentiation of Restrictive From Pericardial Constriction: Assessment of Diastolic Function by

CONSTANTINE N. ARONEY, MBBS, FRACP, TERRENCE D. RUDDY, MD, HUMBERTO DIGHERO, MD, MICHAEL A. FIFER, MD. FACC, CHARLES A. BOUCHER, MD. FACC, IGOR F. PALACIOS. MD, FACC

Diastolic filling variables were studied in 12 patients with constriction was shorter (110 2 14 ms) than in those with the hemodynamic features of constriction, of whom 5 had restrictive cardiomyopathy (195 +- 45 ms, p < 0.01) or in restrictive cardiomyopathy, 5 had pericardial constriction normal subjects (173 + 32 ms, p < 0.01). The percent of and 2 had combined pericardial constriction and restrictive atria1 contribution to left ventricular filling was higher in cardiomyopathy. The values were compared with those in those with restrictive cardiomyopathy (4.5 f 17%) than in 10 normal subjects of comparable age. The filling fractions those with pericardial constriction (21 + 69, p < 0.05) or between 10% and 70% of the diastolic time interval were in normal subjects (24 + 9%, p < 0.01). Peak filling rate in greater in patients with pericardial constriction than in stroke volumes/s, but not in end-diastolic volumes/s, was those with restrictive cardiomyopathy (p < 0.01 between significantly greater in patients with pericardial constric- 20% and SO%, p < 0.05 at lo%, 60% and 70%), with no tion (5.09 2 0.97) than in those with restrictive cardiomy- overlap. The filling fractions in patients with pericardial opathy (3.52 f 0.43, p < 0.01) or in normal subjects constriction were also greater than those in normal subjects (3.98 f 0.70, p < 0.05). between 10% and 60% of the diastolic time interval. The The pattern of left ventricular filling by radionuclide filling fraction was lower in patients with restrictive cardio- angiography helps differentiate restrictive cardiomyopathy myopathy than in normal subjects at 40% of the diastolic from pericardial constriction. time interval (p < 0.05). (J Am Co11Cardiol1989;13:1007-14) The time to peak filling rate in patients with pericardial

Patients presenting with the hemodynamic features of con- (I ,4.5.7-10) however, the hemodynamic features of restric- striction (constrictive physiology*) (1) at cardiac catheter- tive cardiomyopathy may mimic precisely those of restric- ization represent an important diagnostic dilemma in cardi- tive cardiomyopathy. and exploratory thoracotomy may be ology. The differentiation of pericardial constriction from required to examine the for definitive diagnosis restrictive cardiomyopathy in this situation is critical be- (1). Accordingly, a noninvasive method that reliably distin- cause management and prognosis of these two conditions are guishes pericardial constriction from restrictive cardiomyop- different. The separation of such patients on clinical grounds athy would be useful. The pattern of left ventricular filling in is difficult, and noninvasive techniques have been unreliable patients with pericardial constriction may differ from that of (l-6). and, in particular, right ven- patients with restrictive cardiomyopathy, with the constraint tricular can be helpful in establishing the diagnosis on filling occurring earlier in diastole with restrictive cardio- myopathy and later in diastole with pericardial constriction.

From the Cardiac Unit. Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts. Dr. Aroney is an Overseas Clinical Fellow of the National Foundation of Australia. Canberra. Australian Capital Territory, Australia. Dr. Fifer is a Clinician- *Constrictive physiology (II: Equalization of the diastolic pressure of the Scientist of the American Heart Association, Dallas. Texas. v,entricles within 5 mm Hg. elevation of the mean atrial or ventricular diastolic Manuscript received July -75, 1988: revised manuscript received October pressure to ~10 mm Hg. “dip-plateau” pattern of the ventricular pressure 25. 1988, accepted November 17. 1988. curve. prominent Y descent in the right atria1 pressure curve. right ventricular Address for reorints: Igor F. Palacios. MD. Cardiac Unit. Massachusetts end-diastolic pressure 210 of right ventricular systolic pressure and Left General Hospital, Boston. Massachusretts 02114. ventricular ejection fraction zO.40. 1008 ARONEY ET AL. JACC Vol. 13, No. 5 DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY April 1989:1007-14

Table 1. Clinical, Histologic and Hemodynamic Data in 12 Patients

Pt Age (yr)i RVEDP LVEDP Cl No. Gender NYHA RV Biopsy Pericardium (mm Hg) (mm Hg) (literslmin per mj Restrictive Cardiomyopathy

1 40/F III MHIIFIRad Normal 25 29 I so (EchoiThoriNMR) 2 65/M II Normal 11 13 2.14 (Echo) 3 SBIF III Amyloid Normal IO IO 2.40 (Echo) 4 55/F IV MHiIF Normal 16 20 3.90 (Echo/Thor) 5 67iF IV Amyloid Normal 16 16 1.66 (Echo/Thor)

Pericardial Constriction

6 58/M III Normal tThor/Echo II 12 3.20 7 71/M III Normal tThoriEcho/NMR 20 22 2.30 8 62/M III Normal +Thor/Echo/NMR 20 20 1.80 9 52/M IV Normal tThor/Cazf II I1 2.50 IO 74/M IV Normal Echo 30 30 2.20 Combined Constriction/Restrictive

11 52/M IV MHIIFIRF tThor/Ca2’ 12 16 2.00 12 37/F III IFiRad tThor/NMR 19 21 2.56 Ca2+ = calcified pericardium on chest X-ray film; Cl = cardiac index: Echo = two-dimensional (pericardial thickening); F = female: IF = interstitial fibrosis; LVEDP = left ventricular end-diastolic pressure; M = male; MH = myocyte hypertrophy; NMR = nuclear magnetic resonance imaging: NYHA = New York Heart Association functional class; PCWP = pulmonary capillary wedge pressure; RAP = right atrial pressure; Rad = radiation changes; RF = replacement fibrosis; RV = right ventricular; RVEDP = right ventricular end-diastolic pressure: Thor = thoracotomy; t = positive.

We studied 12 patients with constrictive physiology, and fibrillation (Patients 3,9 and 11) with a heart rate at rest < 110 analyzed their diastolic filling patterns at rest, as measured beatslmin. Symptoms were rated according to the New York by radionuclide angiography, to determine whether pericar- Heart Association functional classification for . dial constriction could be differentiated from restrictive The normal group included five men and five women with cardiomyopathy. The results were also compared with those a mean age of 50 2 10 years (range 35 to 64) who had normal of 10 normal subjects of comparable age. findings on physical examination, electrocardiogram (ECG) and rest and exercise first pass radionuclide ventriculogram and no history of . There was no Methods difference in age and heart rate between the constrictive Study patients. Over a 30 month period, all 12 patients physiology and normal groups (Table 2). whose hemodynamic profile met the criteria for constrictive Pericardialconstriction group. The diagnosis of pericar- physiology underwent first pass radionuclide angiography at dial constriction required hemodynamic evidence of con- rest. This group included three patients in whom leg raising strictive physiology, no evidence of a specific form of heart or fluid challenge was performed and diastolic equalization muscle disease on right ventricular endomyocardial biopsy was maintained. Ten normal subjects were also studied. The and pericardial constriction on thoracotomy (Table 1). One constrictive physiology group included five patients with of the five patients (Patient 10) had previous aortic valve pericardial constriction, five with restrictive cardiomyopa- replacement, a normal endomyocardial biopsy and pericar- thy and two with a combination of myocardial fibrosis and dial thickening on echocardiography, but died of sepsis pericardial constriction. The diagnosis in these 12 patients before thoracotomy could be performed. The cause of peri- was made by right ventricular endomyocardial biopsy and cardial constriction in the other four patients was pericardial pericardial evaluation by thoracotomy (Table I). There were constriction after coronary bypass surgery in one, rheuma- seven men and five women with a mean age of 58 + 11 years toid arthritis in one and idiopathic pericardial constriction in (range 37 to 74). Age, gender, diagnosis, result of endomyo- two. Additional corroborative evidence of pericardial dis- cardial biopsy and hemodynamic data are shown in Table 1. ease included the presence of pericardial thickening at Nine of the 12 patients had sinus rhythm, and 3 had atria1 two-dimensional echocardiography, pericardial calcification JACC Vol. 13. No. 5 ARONEY ET AL. 1009 April 1989:1007-14 DLGTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY

Table 2. Standard Left Ventricular Systolic and Filling Indexes in 12 Patients and 10 Normal Subjects p Value

RCM PC Norm RCM v\. PC Norm vs. PC Norm vs. RCM

No. 5 IO Age (yr) 63 i- 9 50 2 II NS NS YS HR (beats/mint 76 ? 6 XI 2 I4 NS NS NS LVEF (%I 53 + 6 66 t x NS 10.05 <0.05 LVEDVI (ml/m’) X0? 17 iI I I6 NS NS NS TTPFR (ms) 110 2 14 I73 2 32

on chest X-ray film and an abnormal thickened pericardium (Baird-Atomic System 77) with the patient at rest. All studies (>4 mm) using magnetic resonance imaging (Table 1) (I I). were performed with the patient upright in the anterior Restrictive cardiomyopathy group. The diagnosis of re- projection. An indwelling catheter was placed in the antecu- strictive cardiomyopathy required hemodynamic evidence bital vein for injection of the radionuclide. Twenty minutes of constrictive physiology and either a specific form of heart after pretreatment with stannous pyrophosphate, 20 mCi of muscle disease demonstrated by endomyocardial biopsy or technetium-99m pertechnetate dissolved <1 ml of normal no pericardial constraint at thoracotomy. A specific form of saline solution was injected and immediately flushed with 20 heart muscle disease was demonstrated in three of the five ml of normal saline solution. Data were accumulated at 25 patients (amyloid heart disease in two and acute myocarditis ms intervals as the radionuclide bolus entered the central in one). Of the remaining two patients, one (Patient I) had circulation. The first transit of the bolus through the major radiation changes including myocyte hypertrophy and inter- vessels and heart chambers was stored on a minicomputer stitial fibrosis secondary to inadequately shielded radiother- system associated with the camera. apy (12) for breast carcinoma, and one (Patient 4) had The data were analyzed with use of the software of the nonspecific myocyte hypertrophy and interstitial fibrosis. Baird-Atomic System 77, as previously described (13-15). No evidence of pericardial constriction was present in these The entire study was completed within 50 s. A region of two patients at thoracotomy. The patient with myocarditis interest was manually placed over the left , and a on biopsy study (Patient 2) had complete symptomatic and time-activity curve generated. Left ventricular peak activity histologic recovery and normal findings on repeat cardiac was considered to be end-diastole, and the activity minimum catheterization after 2 months, precluding a diagnosis of was considered to be end-systole. Only cycles with 270% of pericardial constriction. the maximal end-diastolic activity in the end-diastolic frame Combined pericardial constriction and restrictive cardio- were included for analysis. Background was the activity myopathy. Two patients had mixed features. with both within the left ventricular region of interest before the first pericardial constriction at thoracotomy and an abnormal left ventricular beat. The background-subtracted left ventric- endomyocardial biopsy. Patient 11 had dense pericardial ular beats were summed to generate a single representative adhesions causing marked myocardial scarring evident at cardiac cycle (13,16). There was no bad beat rejection thoracotomy, and large areas of interstitial and replacement because of the few number of cycles included (four to eight). fibrosis on endomyocardial biopsy. Patient 12 had a history Left ventricular ejection fraction was derived from left of mediastinal radiation for Hodgkins disease. In addition to ventricular time-activity curves by the usual method, and pericardial constriction, marked epicardial fibrosis was evi- expressed as a percent of end-diastole. In addition, the dent at thoracotomy. and fibrosis was also present in the end-diastolic outline was analyzed for end-diastolic volume endomyocardial biopsy sample. These patients had patho- with use of a single plane area-length method (13,17). logic features common to both restrictive cardiomyopathy Thefillingfraction curves were calculated by dividing the and pericardial constriction, and were considered separately diastolic time interval for each patient into 10 equal periods from the other groups. and plotting each interval against the simultaneous cumula- Data collection and analysis. First pass radionuclide angi- tive filling fraction ([counts at that time - end-systolic ography was performed with a multicrystal gamma camera counts] x lOO/[end-diastolic counts - end-systolic counts]). 1010 ARONEY ET AL. JACC Vol. 13, No. 5 DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY April 1989:1007-14

RESTRIC J/Vi!- CARDIOMYOPATHY

Figure 1. Raw time-activity curves for a patient (Patient 1) with restrictive cardiomyopathy (left), and a patient (Patient 7) with pericardial constriction (right). * = peak filling rate; AFC = atrial filling contribution; AFP = atrial filling period; TTPFR = time to peak filling rate.

II I I I II I I I I I J/ME

Linear interpolation was performed between points in the Systolic function. Table 2 compares left ventricular ejec- raw data curve to determine filling at each 10% period. This tion fraction and end-diastolic volume index in patients with technique has been used to assess filling using a cineangio- constrictive physiology and normal subjects. There was no graphic method (lo), but has not previously been applied to difference in ejection fraction between patients with pericar- radionuclide data. dial constriction (53 2 6%) and those with restrictive cardio- The peak filling rates were calculated by an algebraic myopathy (48 2 14%). However, normal subjects had a method similar to a method described previously (18), ex- greater ejection fraction (66 ? 8%) than did either patient cept that the derivative of the time-activity curve was group. There was no difference in end-diastolic volume calculated using a wrap around digital filter with coefficients index among groups. (-0.2, -0.1, 0, 0.1, 0.2). This method assumes that the data Diastolic function. Figure 1 demonstrates representative can be described by a third order polynomial function, but raw time-activity curves for restrictive cardiomyopathy (Pa- does not require actual fitting of the data with a polynomial tient 1) and pericardial constriction (Patient 7), and Figure 2 function. The peak filling rates were normalized by end- shows the filling curves for the 10 individual patients nor- diastolic counts and by stroke counts expressed in end- malized to the diastolic filling period. The averaged filling diastolic volumes/s and stroke volumes/s (19). The time to curves normalized to the diastolic filling period for those peak filling rate was calculated as the time at peak filling rate with pericardial constriction or restrictive cardiomyopathy - time at end-systole, and was expressed in milliseconds and for normal subjects are shown in Figure 3. The filling in (ms) (Fig. 1). Time to peak filling rate was also normalized to patients with pericardial constriction was greater than that in the RR interval (time to peak filling rate/RR interval) (20). patients with restrictive cardiomyopathy at all intervals The atrial contribution to filling was derived from time- between 10% and 70% of the diastolic filling interval, with no activity curves, and was defined as the percent of diastolic overlap between groups. This difference was greatest be- filling occurring with atria1 systole. The beginning of atria1 tween 20% and 50% (p < 0.01) of the diastolic filling interval filling was obtained by using the total PR interval from the (peak separation at 40%), with a lesser separation at lo%, electrocardiogram + 40 ms (electromechanical delay) (21). 60% and 70% (p < 0.05). The filling in patients with pericar- The percent of atria1 filling contribution was defined as atria1 dial constriction was greater than that in normal subjects filling contribution x lOO/fillingvolume (Fig. 1). between 10% and 60% of the diastolic time interval (p < 0.01 Statistics. Group data are expressed as mean values f between 10% and 40%, and p < 0.05 at 50% and 60% of the SD. Statistical analysis was done using multivariate one-way diastolic time interval). The filling fraction of patients with analysis of variance, and comparisons of group means was restrictive cardiomyopathy was lower than that of normal by Newman-Keuls test (22). The differences were consid- subjects at 40% of the diastolic time interval (p < 0.05); ered significant at the level of p < 0.05. therefore, at 40% of this interval, there was significant separation of cumulative filling fractions in all three groups. Flattening of the filling curve in normal subjects (diastasis) Results occurred at about 75% of the diastolic time interval, whereas Table 1 compares the clinical, histologic and hemody- in those with pericardial constriction (where it represents the namic data of the patient groups. cessation of rapid filling caused by the constraints of the JACC Vol. 13. No. 5 ARONEY ET AL. 101 I April 198Y:lOO7-14 DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY

Figure 2. Left panel depicts normalized diastolic filling curves for five patients with restrictive cardio- myopathy. Solid triangle = sinus rhythm: open tri- angle = atria1 fibrillation. Right panel depicts normal- ized diastolic filling curves for five patients with pericardial constriction. Solid square = sinus rhythm: open square = atria1 fibrillation.

PERCENT DURATION OF DIASTOLE

pericardium), it occurred earlier at about 50% of the diastolic trend to a higher rate in patients with pericardial constriction time interval (Fig. 3). (2.5 t 0.5) than in those with restrictive cardiomyopathy Table 2 and Figure 4 sho\il the results of other variables (1.90 + 0.441, and no significant difference from normal used to assess left ventricular_filling. The time to peak filling subjects (2.65 2 0.63). However, when the peak filling rate rate in those with pericardial constriction (110 ? 14 ms) was was normalized by stroke volume, it was significantly higher shorter than in those with restrictive cardiomyopathy (195 * in those with pericardial constriction (5.09 F 0.97) than in 45 ms, p < 0.01) and in normal subjects (173 2 32 ms. p < those with restrictive cardiomyopathy (3.52 I 0.43, p < O.Ol), and this difference was maintained after correction for 0.01) and in normal subjects (3.98 t 0.70. p < 0.05). the RR interval. There was no overlap between time to peak Thus, the extent of early filling was consistently increased filling rate for patients with pericardial constriction (100 to in patients Lzith pericardial constriction, with no overlap 125 ms) and restrictive cardiomyopathy (175 to 275 ms). The between those with pericardial constriction and those with atria1 contribution to filling was greater in those with restric- restrictive cardiomyopathy from 10% to 70% of the diastolic tive cardiomyopathy (45 _t 17’S, range 33% to 70%) than in time interval. In addition, in patients with pericardial con- those with pericardial constriction (21 t 6%, range 12% to striction, the time to peak filling occurred earlier and the 27%, p < 0.05) or in normal subjects (24 it 9%. p < 0.01). peak filling rate (in stroke volumes/s) was greater than in The peak filling rate normalized by end-diastolic volume was patients with restrictive cardiomyopathy or in normal sub- not different among the three groups, although there was a jects. Combined restrictive cardiomyopathy and pericardial con- striction. Two patients had both myocardial fibrosis on Figure 3. Averaged diastolic filling curves for 5 patients with endomyocardial biopsy and pericardial constriction at tho- pericardialconstriction (square), 5 patients with restrictive cardio- racotomy (Patients I1 and 12, Table I), including one (Pa- myopathy (triangle) and 10 normal subjects (circle). Filling values are + SD. tient 11) with pericardial calcification. In Patient 11, the filling curve was similar to that in those with restrictive cardiomyopathy, suggesting that the dominant component contributing to reduced filling in this patient was restrictive cardiomyopathy. In Patient 12, the extent of early filling was intermediate compared with that in patients with pericardial constriction or restrictive cardiomyopathy.

Discussion The differentiation of pericardial constriction from re- strictive cardiomyopathy is very difficult and often relies on thoracotomy for definitive exclusion of pericardial constric- tion. The risks of such surgery in patients with restrictive cardiomyopathy, particularly those with cardiac amyloido- FFRCENT DURA T/ON OF D/AS JOL E sis, have been well documented (23,24). Pericardial stripping 1012 ARONEY ET AL. JACC Vol. 13. No. S DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY April 1989: 1007-14

TIME TO PEAK PEAK FILLING RATE ATRIAL FILLING FILLING RATE (msl (W/s) CONTRIBUTION(%)

4o07 g,“l

. Figure4. Valuesfor time to peak filling rate (ms), peak 6.0 filling rate (stroke volume [SV]/s) and atria1 filling con- L . tribution (9%)are shown for each group, including mean 35 + SD. C = pericardial constriction; NORM = normal t subjects; R-CM = restrictive cardiomyopathy.

15 t

I L 1 I NORMC R-CM NORMC R-CM

also carries a mortality risk, which could be reduced if it diography and plotted against the fractionated diastolic time were possible to identify that subset of patients with con- interval (10,30) have provided the most insight in character- comitant pericardial constriction and restrictive cardiomy- izing left ventricular filling in pericardial constriction and opathy who may not benefit from surgery. restrictive cardiomyopathy. These techniques support our Limitations of other techniques. Endomyocardial biopsy findings of divergent early filling patterns in pericardial (25,26) may confirm the presence of specific infiltrative constriction and restrictive cardiomyopathy. However, they disease, provide prognostic information and even suggest call for geometric assumptions about the shape of the left therapy. A normal biopsy examination excludes restrictive ventricle, and require frame by frame tracing of multiple left cardiomyopathy (25). However, a biopsy showing nonspe- ventricular silhouettes, which precludes their use in routine cific changes such as hypertrophy and interstitial or replace- clinical practice. Furthermore, regional differences in seg- ment fibrosis, or both, does not exclude concomitant peri- mental wall thickening, excursion and rates of change are cardial constriction (Patients 11 and 12) or identify whether important limitations of these techniques (39,40). The this or restrictive cardiomyopathy is the major determinant present study applied this concept to radionuclide methods, of constrictive physiology (25,27). Echocardiography may thereby avoiding geometric assumptions and planimetry. It allow for qualitative discrimination of normal from thickened also showed, with use of a standard noninvasive method, a pericardium (28) but does not quantify pericardial thickness clear separation of patients with pericardial constriction and accurately (29,30). Abnormal septal motion (31-33), flat- restrictive cardiomyopathy on the basis of their diastolic tening of left ventricular posterior wall diastolic movement filling curves. There was no overlap of patients with pericar- (29,31,33,34) and increased left ventricular wall thickness dial constriction and those with restrictive cardiomyopathy (2,34,35) have been described in restrictive cardiomyopathy. between 10% and 70% of the diastolic time interval, and of Computed tomographic scanning has been proposed to iden- patients with pericardial constriction and normal subjects tify pericardial thickening and tumor deposition (36,37). between 10% and 60% of the diastolic time interval. Nuclear magnetic resonance imaging appears to be an accu- Restrictive cardiomyopathy and pericardial constriction rate technique for assessing pericardial thickness, and a may coexist. In both our patients with combined disease, thickness >4 mm has been considered abnormal (11). severe myocardial involvement was present, and in one Hemodynamic data have not proved to be reliable in (Patient II), a restrictive pattern of filling dominated and differentiating pericardial constriction from restrictive car- concealed the coexistence of pericardial constriction. More diomyopathy in patients with constrictive physiology (3- information is required in patients with combined pericardial 6,10,30,38). Patients with restrictive cardiomyopathy with constriction and restrictive cardiomyopathy, but this tech- both equal (constrictive physiology) and unequal ventricular nique may 1) identify severe associated restrictive cardio- filling pressures have been assessed as a homogeneous group myopathy, which may reduce the chances of a successful (lo), whereas other studies of restrictive cardiomyopathy surgical outcome; or 2) identify pericardial constriction as have not reported individual hemodynamic data (6,30). the major determinant of reduced filling, suggesting that Therefore, a disparate variety of abnormalities of diastolic pericardial stripping will be beneficial. function have been described in restrictive cardiomyopathy. Traditional radionuclide filling indexes. The differences in Diastolic filling curves. Left ventricular filling curves de- filling curves were corroborated by a significant shortening rived from the fraction of the diastolic filling volume from of the time to peak filling rate in patients with pericardial left ventricular angiography or digitized M-mode echocar- constriction compared with that in normal subjects and JACC Vol. 13, No. S ARONEY ET AL. 1013 April 1989:1007-14 DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY

patients with restrictive cardiomyopathy. This shortening high time resolution and rejection of bad beats, probably was maintained even when time to peak filling rate was using a list mode acquisition with reformatting (21). corrected for RR interval (20). The percent of atria1 filling This study includes patients with atrialfihrillation, whose contribution, a strong inverse correlate of early filling (21). is abnormal filling characteristics may have confounded the increased in patients with restrictive cardiomyopathy com- influence of pericardial constriction or restrictive cardiomy- pared with those with pericardial constriction and normal opathy on left ventricular diastolic function. However, our subjects, confirming the marked reduction in early filling in three patients with constrictive physiology and atria1 fibril- restrictive cardiomyopathy and a greater reliance on atrial lation had filling characteristics similar to those of patients in systole. The percent atria1 filling contribution may be more sinus rhythm (Fig. 2) because they did not have unusual sensitive than the diastolic filling curves in separating pa- heart rates or wide variations in RR intervals. Information is tients with restrictive cardiomyopathy from normal subjects required to determine the filling pattern in patients with because the significance level was greater (p < 0.01 without restrictive cardiomyopathy who do not have constrictive overlap) than with diastolic filling fraction at 40% of the physiology, although their differentiation from those with diastolic time interval (p < 0.05). However, atria1 fibrillation pericardial constriction on hemodynamic grounds is clear. may occur with restrictive cardiomyopathy, precluding the Conclusions. First pass radionuclide angiography in pa- use of percent atria1 filling contribution. tients with constrictive physiology allows differentiation of pericardial constriction from restrictive cardiomyopathy. The peak rate of ventricular filling has been shown to he The extent of early filling from 10% to 70% of diastole is preload dependent (411, and approaches to normalizing this index have included dividing by instantaneous or end- greater in pericardial constriction than in restrictive cardio- diastolic volume. These techniques adjust for variation in myopathy. This observation is confirmed by the atria1 con- ventricular size, but will conceal elevations in filling when tribution to filling, the time to peak filling rate and the peak filling rate (stroke volumes/s), each of which separates the there are subtle elevations in end-diastolic volume (as with patient groups without overlap. The demonstration of early the patients with pericardial constriction in this study). rapid filling with this technique indicates that pericardial Angiographically determined filling rate has been shown to constriction is likely. Reduced or normal early filling is correlate with stroke volume in normal patients (42) which demonstrated in patients with restrictive cardiomyopathy or suggests normalization of filling rate by dividing by stroke with combined disease. We suggest that patients with con- volume. Just as we employed the fraction of filling volume strictive physiology syndrome could be evaluated by a for describing the pattern of filling, so we employed filling combination of endomyocardial biopsy and first pass radio- volume to normalize peak filling rate (19). This variable nuclide angiography to differentiate pericardial constriction distinguished pericardial constriction from restrictive cardio- from restrictive cardiomyopathy, and to select candidates myopathy without overlap, whereas peak filling rate cor- who may benefit from thoracotomy and . rected by end-diastolic volume did not. Limitations of the study. All patients with pericardial constriction and restrictive cardiomyopathy in this study had References overt and symptomatic disease, with a moderate or marked I Johnson PA, Palacios IF. Non-dilated . Adv Intern Med elevation and equalization of ventricular filling pressures. 1984:30:243-74. This technique has not been validated in patients with occult 2. Borer JS, Henry WL, Epstein SE. Echocardiographic observations in disease, who do not have elevation of ventricular filling patients with systemic infiltrative disease involving the heart. Am J pressure at rest. Although the number of patients studied is Cardiol 1979;39: 184-8. small, the differences are statistically significant. We con- 3. Gunnar RM, Dillon RF, Wallyn RJ, Elisberg EL. The physiologic and clinical similarity between primary amyloid of the heart and constrictive sider that the lack of overlap of filling curves, time to peak pericarditis. Circulation 1953:2:827-32. filling rate, peak filling rate and percent atria1 filling contri- 4. Meaney E. Shabetai R, Bhargavia V, et al. , constric- bution indicates the merit of this technique. The usefulness tive pericarditis and restrictive cardiomyopathy. Am J Cardiol 1976;38: of peak filling rate may be reduced by its preload dependence S47-56. (41), although this index provided a clear separation of 5. Swanton RH, Brooksby IAB, Davies MJ. Coltart DJ, Jenkins BS. Webb-Peoploe M. Systolic and diastolic ventricular function in cardiac groups when expressed as stroke volumes/s. amyloidosis: studies in six cases diagnosed with endomyocardial biopsy. This study was performeti with first pass radionuclide Am J Cardiol 1977;39:658&l. angiography, which ensures that adequate counts are ob- 6. Janos GG, Arjunan K. Meyer RA, Engel P. Kaplan S. Differentiation of tained with few beats and without the need for gating. The constrictive pericarditis and restrictive cardiomyopathy using digitized echocardiography. J Am Coil Cardiol 1983;1:541-9. data are acquired in list mode and then reformatted, taking 7. Hoyninger-Huene CBJ. Systemic amyloidosis presenting as constrictive every end-diastole with ~70% of peak activity. These con- pericarditis: a case studied with cardiac catheterization. Am Heart J cepts could theoretically be applied to R wave-gated blood 1964:67:29&l. pool scanning, but this procedure would require high counts, 8. Bridgen W. Cardiac amyloidosis. Prog Cardiovasc Dis 196437:142-50. 1014 ARONEY ET AL. JACC Vol. 13, No. 5 DIASTOLIC FILLING BY RADIONUCLIDE ANGIOGRAPHY April 1989:1007-14

9. Schroeder JS, Billingham ME, Rider AK. Cardiac amyloidosis. Diagnosis 27. Seifert FC, Miller DC, Oesterle SN, Oyer PE, Stinson EB, Shumway NE. by transvenous endomyocardial biopsy. Am J Med 1975;59:26%73. Surgical treatment of constrictive pericarditis: analysis of outcome and 10. Tyberg TH, Goodyear AV, Hurst VW, Alexander J, Langou RA. Left diagnostic error. Circulation 1985;72(supplH):lI-26&73. ventricular filling in differentiating restrictive amyloid cardiomyopathy 28. Schnittger I, Bowden RE, Abrams J, Popp RL. Echocardiography: and constrictive pericarditis. Am J Cardiol 1981;47:791-6. pericardial thickening and constrictive pericarditis. Am J Cardiol1978;42: 11. Sechtem V, Tscholakoff D, Higgins CB. MRI of the pericardium. AJR 388-95. 1986;146:245-52. 29. Voelkel AG, Pietro DA, Folland ED, Fisher ML, Parisi AF. Echocardio- 12. Cohn KE, Stewart JR, Fajardo LF, Hancock EW. Heart disease follow- graphic features of constrictive pericarditis. Circulation 1978;58:871-5. ing radiation. Medicine 1967;46:281-98. 30. Pandian NG, Skorton DJ, Kieso RA, Kerber RE. Diagnosis of constric- 13. Boucher CA, Anderson MD, Schneider MS, Murphy JH, Okada RD, tive pericarditis by two-dimensional echocardiography: studies in a new Kanarek DJ. Left ventricular function before and after reaching the experimental model and in patients. J Am Coll Cardiol 1984;4:1164-73. anaerobic threshold. Chest 1985;87:145-50. 31. Gibson TC, Grossman W, McLaurin LP, Moos S, Craige E. An echocar- 14. Miller DD, Ruddy TD. Zusman RM, et al. Left ventricular ejection diographic study of the in constrictive pericar- fraction response during exercise in asymptomatic systemic hyperten- ditis. Br Heart J 1976;38:738-43. sion Am J Cardiol 1987;59:408-13. 32. Benotti JR, Grossman W, Cohn PF. Clinical profile of restrictive cardio- 15. Salahi F, Systrom DM, DiCola VC, et al. Cardiopulmonary exercise myopathy. Circulation 1980;61:120612. testing in DDD compared to VVI pacing modes. Am J Noninvas Cardiol 1987;1:244-50. 33. Engel PJ. Fowler NO, Tei CW. et al. M-mode echocardiography in 16. Scholz PM, Rerych SK, Moran JF, et al. Quantitative radionuclide constrictive pericarditis. J Am Coll Cardiol 1985;6:4714. angiography. Cathet Cardiovasc Diagn 1980;6:265-83. 34. Child JS, Levisman JA, Abbasi AS, MacAlpin RN. Echocardiographic 17. Anderson PAW, Rerych SK, Moore TE, Jones RH. Accuracy of left manifestations of infiltrative cardiomyopathy: a report of seven cases due ventricular end-diastolic dimension obtained by radionuclide angiogra- to amyloid. Chest 1976;70:72631. phy. J Nucl Med 1981;22:500-5. 35. St John Sutton MG, Reichek N, Kastor JA, Giuliani ER. Computerized 18. Magorien DJ, Shaffer P, Bush CA, et al. Assessment of left ventricular M-mode echocardiographic analysis of left ventricular dysfunction in pressure-volume relations using gated radionuclide angiography, echocar- cardiac amyloid. Circulation 1982;66:790-9. diography, and micromanometry pressure recordings. Circulation 1983; 36. Moncada R, Kotler MN, Churchill RJ, Demos TC, Jacobs W, Steiner 67384453. RM. Multimodality approach to pericardial imaging. Cardiovasc Clin 19. Fifer MA, Borow KM, Colan SD, Lorell BH. Early diastolic left 1986;17:409-41. ventricular function in children and adults with aortic stenosis. J Am Coll Cardiol 1985;5:1147-54. 37. Isner MI, Carter BL, Bankoff MS, et al. Differentiation of constrictive pericarditis from restrictive cardiomyopathy by computed tomographic 20. Mancini GBJ, Slutsky RA, Norris SL, Bhargava V, Ashburn WL, Higgins imaging. Am Heart J 1983;105:1019-25. CB. Radionuclide analysis of peak filling rate, filling fraction, and time to peak filling rate. Am J Cardiol 1983;51:43-5 I. 38. Hetzel PS. Wood EH. Burchell HP. Pressure pulses in the right side of the heart in a case of amyloid disease and in a case of idiopathic heart failure 21. Bonow RO, Frederick TM, Bacharach SL, et al. Atrial systole and left simulating constrictive pericarditis. Mayo Clin Proc 1953;28:107-12. ventricular filling in hypertrophic cardiomyopathy: effect of verapamil. Am J Cardiol 1983;51:1386-91. 39. Pandian NG, Skorton DJ, Collins SM, Falsetti HL, Burke ER, Kerber 22. Godfrey K. Comparing the means of several groups. N Engl J Med RE. Heterogeneity of left ventricular segmental wall thickening and 1985;313:1450-6. excursion in 2-dimensional echocardiography of normal human subjects. Am J Cardiol 1983;51:1667-73. 23. Couter WT, Reichert RE Jr. Primary systemic amyloid mimickingchronic constrictive pericardial disease. Circulation 1950;2:441-6. 40. Pollick C, Fitzgerald PJ, Popp RL. Variability of digitized echocardiog- raphy: size, source, and means of reduction. Am J Cardiol 1983;51:576 24. Case records of the Massachusetts General Hospital (case 48-1972). N 82. Engl J Med 1972;287:1138-45. 25. Schoenfeld MH, Supple EW, Dee GWD Jr, Fallon JT, Palacios IF. 41, Choong CY, Herrmann HC, Weyman AE, Fifer MA. Preload dependence Restrictive cardiomyopathy versus constrictive pericarditis: role of en- of Doppler-derived indexes of left ventricular diastolic function in hu- domyocardial biopsy in avoiding unnecessary thoracotomy. Circulation mans. J Am Coll Cardiol 1987;10:800-8. 1987;75:1012-7. 42. Hammermeister KE, Warbasse JR. The rate of change of left ventricular 26. Schroeder J, Billingham ME, Rider A. Cardiac amyloidosis: diagnosis by volume in man. 11. Diastolic events in health and disease. Circulation transvenous endomyocardial biopsy. Am J Med 1975;59:269-73. 1974;49:739-47.