- Email: [email protected]
Antimyosin antibody imaging in experimental aortic regurgitation
Antimyosin antibody imaging in experimental aortic regurgitation Ping Lu, MD, Pat Zanzonico, PhD, Steven M. Goldfine, PhD, Ruth Hardoff, MD, DSc, Norman Magid, MD, Raffaele Gentile, MD, Edmund M. Herrold, MD, PhD, and Jeffrey S. Borer, MD, with the technical assistance of Mark Morgan, BS Background. Fiber dropout and myocyte necrosis precede heart failure in experimental aortic regurgitation (AR). The current study aimed to determine whether this process can be detected by noninvasive scintigraphic imaging. Methods and Results. 111In-labeled antimyosin antibody Fab fragment (1 to 1.5 mCi) (Myoscint) was administered to each of 34 New Zealand White rabbits: 11 early (3 to 5 weeks) after surgical AR induction; 9 late (98 to 128 weeks) after AR induction; 5 normal and 3 sham-operated age-matched with early AR; and 3 normal and 3 sham-operated age-matched with late AR. Echocardiographic fractional shortening was indistinguishable among control, early AR, and late AR groups. In vivo gamma camera imaging 24 and 48 hours after isotope administration, post-mortem heart activity determination (percentage injected dose per gram), and autoradiography were performed. At 24 and 48 hours, heart-to-lung counts-per-pixel ratios from in vivo images were greater (p < 0.05) in the late AR rabbits than in each of the three other groups. No significant differences were found when early AR and older or younger control rabbits were compared. Heart activity (percentage injected dose per gram) in late AR rabbits trended toward higher values than in age-matched control rabbits (p = 0.057), but in early AR it was indistinguishable from that in the corresponding control (p = 0.413, difference not significant). The autoradiographic endocardial/epicardial activity ratio in late AR rabbits was greater than in control and early AR rabbits (1.27 +_ 0.13 vs 1.06 __ 0.09 and vs 1.13 -- 0.10, respectively, p < 0.02). Conclusions. Whereas isotope uptake in late AR rabbits differed from that in control and early AR rabbits, systolic function was indistinguishable. Thus 111In-labeled antimyosin antibody imaging may permit noninvasive detection of AR-induced myocardial damage before functional deterioration. (J Nucl Cardiol 1997;4:25-32.) Key Words: volume overload • aortic regurgitation- cardiac imaging • antimyosin antibody
From the Divisionof Cardiologyand the Divisionof NuclearMedicine, Cornell University Medical College, The New York HospitalCornell Medical Center, New York, N.Y. Supported in part by grant 1R29HL-40679-01, National Heart, Lung and Blood Institute (Bethesda, Md.); by grants from McNeil Pharmaceutical (Raritan, N.J.), Centocor (Malvern, Pa.) (which provided antimyosin antibody), The Howard Gilman Foundation (New York, N.Y.), The Daniel and Elaine Sargent CharitableTrust (New York, N.Y.), The David Margolis Foundation (New York, N.Y.), and The Charles and Jean Brunie Foundation (Bronxville, N.Y.); and by gifts from Rorlald and Jean Schiavone, Mary Jane Voute Arrigoni, and the late William Voute. During their participationin this study, Dr. Lu was a Howard Gilman FoundationAdvancedVisiting Fellow in CardiovascularResearch, Dr. Hardoff was a Howard Gilman Foundation Distinguished Visiting Scientist, and Dr. Gentile was a Howard Gilman Foundation Visiting Scientist at Cornell UniversityMedical College. Reprint requests: Jeffrey S. Borer, MD, The New York HospitalCornell Medical Center, 525 East 68th St., New York, NY 10021. Copyright © 1997 by American Society of Nuclear Cardiology. 1071-3581/97/$5.00+0 43/1/76175
Myocyte dropout and myocardial fibrosis have been observed in patients with aortic regurgitation (AR) in whom congestive heart failure develops and who undergo valve replacement. 1-8 These pathologic alterations are thought to contribute to the development of heart failure. 9-11 Similar histopathologic abnormalities have been reported in preliminary form in an experimental model of AR that closely simulates the characteristics and natural history of the human disease/2 Myocyte necrosis and fiber dropout and fibrosis were seen in most animals with chronic AR but were most prevalent in animals with congestive heart failure and were less marked or were absent among hemodynamically compensated animals. These observations suggest that in this model and perhaps in human beings, if fibrosis or myocyte necrosis could be identified noninvasively, the subsequent development of heart failure might be predictable. 25
26
Lu et al. Noninvasiveimaging in aortic regurgitation
Normally myosin is found entirely within the myocardial cell. Myocyte necrosis exposes myosin to the extracellular milieu. Therefore myocyte necrosis might be detected by external gamma camera imaging if radiolabeled antibody directed at myosin were administered. This technique has been used successfully to identify and localize myocyte necrosis due to myocardial infarction, myocarditis, and cardiomyopathy. 13-31 On the basis of these observations, we used antimyosin antibody imaging in animals with chronic severe experimental AR and in appropriate control animals to test the hypothesis that this noninvasive imaging technique can detect necrosis in myocardium subjected to prolonged volume loading in AR.
METHODS Animal Population. Thirty-four New Zealand White rabbits were studied. In 20 of the 34 rabbits, AR was surgically induced using closed-chest modification32 of our earlier procedures, 33'34 a s described briefly below. Of these 20, 9 (age 32.3 +_ 8.2 months, weight 3.80 +_ 0.30 kg) had moderate to severe AR (regurgitant fraction [RF] 56.6% + 7.2%) and were studied in the chronic phase, 98 to 128 weeks after AR induction. These rabbits did not show clinical signs of heart failure, and fractional shortening was 33.5 -+ 2.6%, within the limits we have previously defined as normal?3 The remaining 11 animals (age 6.4 _+ 0.9 months, weight 3.05 -+ 0.27 kg), with AR comparable in severity to that of the chronically regurgitant animals (RF 50.8 _+ 11.9% and fractional shortening 32.6 -+ 5.2%, no significant difference vs chronic AR), were studied 3 to 5 weeks after AR induction. This group was studied to enable accurate interpretation of results in animals with chronic AR by permitting determination of whether myocyte damage occurs early after AR development. Negative controls included normal uninstrumented animals and shamoperated rabbits prepared identically to AR animals except that the aortic valve was not perforated. Because there were significant differences in age and weight between the early and late AR groups, control animals were matched in age with each group. The younger control group (age 5.6 _+ 0.8 months, weight 2.95 + 0.27 kg) comprised 5 normal and 3 shamoperated rabbits; the older control group (age 33.7 -+ 8.9 months, weight 3.74 -+ 0.32 kg) included 3 normal and 3 sham-operated rabbits. Echocardiographic characteristics of AR and control rabbits are presented in Table 1. AR Induction. Rabbits were anesthetized with subcutaneous injection of 35 mg/kg ketamine and 5 mg/kg xylazine, with supplementary ketamine as needed during operation. The carotid artery was exposed and isolated. A 5F catheter was placed in the proximal carotid to monitor blood pressure and to act as a sheath for the introduction of a bevel-tipped catheter. The second catheter was advanced retrograde into the left ventricle through an aortic valve cusp during continuous arterial blood pressure monitoring. After the aortic valve was adequately perforated as determined by Doppler echocardiographic RF assessment (as discussed below) the catheter was
Journal of Nuclear Cardiology January/Febrnary1997, Part 1
withdrawn, the carotid ligated, and the incision closed in layers.31-35 Echocardiography. After induction of AR, aortic and pulmonary artery flows were measured by Doppler echocardiography to quantify RF, as previously described. 31-34 Twodimensional, M-mode, and Doppler echocardiograms were performed within 3 days after AR induction and once per month thereafter until within 1 week before antimyosin antibody imaging, according to methods previously described and validated. 3z-35Animals were sedated (35 mg/kg ketamine and 5 mg/kg xylazine); they were placed in left lateral decubitus; electrocardiographic leads were attached, and two-dimensionally directed M-mode echocardiograms of the left ventricle were obtained with a Doppler-echocardiographic system (Hewlett-Packard, Sonos 500 [Revision]) equipped with a 5-mHz, short-focus transducer and recorded on paper at 100 mm/sec. Fractional shortening was calculated, and left ventricular mass was determined as previously described and validated against autopsy measurements?T M Radionuclide Imaging. A Fab fragment of antimyosin monoclonal antibody coupled to diethylenetriaminepentaacetic acid (DTPA) (Myoscint, Centocor, Malvern, Pa.) was labeled with l llIn according to previously described methods. 13 Each rabbit received 500 Ixg of antibody labeled with 1 to 1.5 mCi l 1~in in 2.5 ml normal saline by bolus injection into an ear vein. Gamma camera images were acquired at 24 and 48 hours after antibody administration. Rabbits were sedated with intramuscular injection of 35 mg/kg ketamine and 5 mg/kg xylazine. With the rabbit placed and gently restrained in the supine position, chest images were obtained in the anterior and 45-degree left anterior oblique positions with a gamma camera (GE StarPort 400 AC) with 3.5 mm pinhole collimator. In addition, whole-body and fourfold-zoomed anterior and 45degree left anterior oblique images were acquired with a medium-energy parallel-hole collimator (GE StarCam 500A). In both cases, dual-photon acquisition of 300,000 to 500,000 counts was performed with photopeak energy windows of 173 keV -+ 15% and 247 keV -+ 15% and a 256 × 256 image matrix. To ensure comparability of technical factors associated with imaging, animals were paired or combined in small groups for isotope administration and imaging at the same session: in each pair or small group an AR animal was matched with at least one normal or sham-operated control animal. Image Analysis. Of the several images acquired, those obtained in the anterior position with the pinhole collimator exhibited the most pronounced myocardial "targeting" and heart-to-lung and heart-to-liver mean counts-per-pixel ratios. Therefore, these unprocessed images were used to quantitate antibody uptake and localization in the heart. A cardiac region of interest was defined and contained both ventricles. Regions of interest over the lung and liver were defined. Mean counts per pixel, were determined for each region, and mean heart-tolung and heart-to-liver counts-per-pixel ratios were calculated for the 24- and 48-hour images, as described by investigators. 2~,28,3° Interobserver count ratio variation was defined by two independent and blinded obseIvers. Intraobserver variation was calculated from consecutive determinations by a single
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
27
Table 1. E c h o c a r d i o g r a p h i c a n d a g e c h a r a c t e r i s t i c s o f A R a n d c o n t r o l rabbits at t i m e o f a n t i m y o s i n antibody imaging
Younger control
Early AR
Older control
Late AR
Age (mo)
Weight (kg)
l e f t ventricle internal diameter, diastole (cm)
Posterior wall thickness, diastole (cm)
Fractional shortening
5.6 _+ 0.8 (5.0-7.0) n=8 6.4 +_ 0.9 (5.0-7.8) n = II 33.7 _+ 8.9* (22.0-50.0) n=6 32.3 _+ 8.2"~ (20.0-50.0) n=9
2.95 -_ 0.27 (2.70-3.63) n=8 3.05 _+ 0.27 (2.66-3.60) n = II 3.74 _+ 0.32* (3.34-4.25) n=6 3.80 _+ 0 . 3 0 t (3.42-4.44) n=9
1.44 +_ 0.16 (1.20-1.67) n=7 1.88 +_ 0.34 (1.30-2.39) n = II 1.38 _+ 0.15 (1.20- 1.50) n=4 2.07 _+ 0.28:~ ( 1.60-2.50) n=9
0.23 _+ 0.06 (0.15-0.32) n=7 0.23 _+ 0.04 (0.18-0.31) n = II 0.26 + 0.03 (0.23-0.31 ) n=4 0.25 + 0.04 (0.20-0.34) n=9
32.50 _+ 5.10 (27.5-40.0) n=7 32.55 _+ 5.18 (23.1-37.5) n = II 36.00 _+ 7.07 (27.0-44.0) n=4 33.51 _+ 2.63 (31.6-39.0) n=9
(%)
*p < 0.02 vs younger control. t P < 0.001 vs early AIL ~p < 0.00l vs older control.
observer who was blinded to the identity of the rabbit being evaluated.
Autoradiography and Tissue Biodistribution. Rabbits were systemically heparinized (200 U/kg) and killed by intravenous injection of 100 mmol/L cadmium chloride immediately after the acquisition of the 48-hour images and while still anesthetized. In some experiments, catheters were placed in both the carotid artery and jugular vein so that the heart could be perfusion-fixed in situ, whereas in others the heart was immediately removed from the chest and perfused with a catheter placed directly into the root of the aorta. In either case, the heart was initially flushed with heparinized (10 U/ml) 100 mmol/L phosphate-buffered saline (pH 7.2) and subsequently fixed with either phosphate-buffered 4% paraformaldehyde or 2% glutaraldehyde and 2% paraformaldehyde. In general, no blood was apparent on the surface of the heart after either procedure. The fixed heart was cut into 3 to 4 mm slices from apex to base with a razor blade or a meat slicer. Alternating slices were used to determine radioactivity concentration in the heart and 1~~In antimyosin antibody distribution within the heart. Slices were placed in either 4% formalin or 15% sucrose-phosphatebuffered saline. Slices in 4% formalin were weighed and counted in an automated scintillation well counter calibrated for 111in" Net count rates were corrected for radioactive decay and converted to percentage of administered radioactivity per gram. Slices placed in sucrose were incubated at 4 ° C for -->3 hours, embedded in OTC medium (Tissue-Tek), frozen in isopentane at liquid nitrogen temperatures, and cut into 20 ixm-thick sections with a Haaker-Bright (Instrument Company, Huntingdon, England) cryostat to determine 111In antimyosin antibody localization autoradiographically. Approximately every tenth cryosection was exposed to scientific imaging film
(Kodak XO Mat AR) at - 7 0 ° C for 7 to 21 days. Intensifier screens were used to optimize images for analysis. As noted above, AR and control animals were matched for isotope administration and imaging; all members of these matched pairs or small groups were killed the same day and were placed on the same autoradiographic film to ensure identical exposure times for matched AR and control rabbits and thus for the four subgroups of AR and control animals. Autoradiograms were magnified and recorded with a high-resolution charge-coupled device camera (Sierra Scientific). The charge-coupled device image then was divided manually into endocardial and epicardial halves. The optical densities of the endocardial and epicardial regions were determined (MCID software version 4.2, Imaging Research). The ratios of endocardial-to-epicardial optical density were averaged for each rabbit, and the resulting values were used to calculate group averages for late AR, early AR, older control, and younger control rabbits. Statistical Analysis. Nonparametric Kruskal-Wallis one-way analysis of variance was used to compare heart-tolung and heart-to-liver counts-per-pixel ratios from in vivo images and to compare tissue concentrations of the heart and the endocardial-to-epicardial optical density ratios from autoradiograms. Late AR, early AR, older control, and younger control rabbits were compared.
RESULTS E c h o c a r d i o g r a p h i c f i n d i n g s . Table 1 presents the e c h o c a r d i o g r a p h i c findings in late AR, early A R , and older and y o u n g e r control animals. N o significant differ-
28
Lu et al. Noninvasive imaging in aortic regurgitation
48
Journal of Nuclear Cardiology January/February 1997, Part 1
I~r [
_ e, " t e.
qR
Figure 1. Anterior pinhole-collimator gamma camera images of thorax of younger control (normal), older control (shamoperated), early AR (acute AR), and late AR (chronic AR) rabbits 48 hours after injection of l~lIn-!abeled antimyosin antibody Fab fragment. At 48 hours, heart (center of each image) is better defined and reveals more intense isotope uptake relative to surrounding tissues (e.g., lungs and liver) in late AR rabbit than in early AR rabbit, younger control, or older control rabbit.
ences in fractional shortening were seen among the four groups. in Vivo Imaging. At 24 and at 48 hours after isotope administration, in vivo images revealed diffuse cardiac activity in all four groups (late AR, early AR, older control, and younger control) (Figure 1), except in one late AR rabbit that manifested marked inferoapical uptake (Figure 2). Heart-to-lung and heart-to-liver counts-per-pixel ratios were significantly greater in the late AR rabbits than in each of the three other groups at 24-hour imaging (Table 2). No significant differences were found among the three remaining groups, although younger control animals tended to exhibit a cardiac uptake lower than that of early AR rabbits. The heartto-lung ratios also were greater in late AR rabbits than in each of the three other groups at 48-hour imaging; for the heart-to-liver ratio, statistical significance was retained at 48 hours only for late AR versus older control animals (Table 2). Interobserver and intraobserver correlations were relatively strong for in vivo gamma camera measures of the heart, lung, and liver uptake: correlation coefficients were 0.997, 0.794, and 0.972 (all p < 0.001), respectively, between two blinded independent observers, and
were 0.98, 0.88, and 0.83 (allp < 0.01), respectively, for two consecutive determinations by a single observer. Autoradiographic Distribution. The imaging results were mirrored in the autoradiographic findings. Intense uptake was seen only in late AR animals (Figure 3). Moreover, radiotracer uptake was most prominent in the subendocardium of animals with late AR; such regional sequestration was not seen in normal or shamoperated animals of either age group or in early AR animals. In late AR rabbits, the average endocardial-toepicardial uptake ratio was 1.27 _+ 0.13 vs 1.06 + 0.09 in older controls (p < 0.02). Late AR rabbits also manifested significantly greater absolute subendocardial uptake than did age-matched control rabbits (p = 0.02) (Table 3). Ratios in early AR and younger control animals were 1.13 _+ 0.10 and 1.05 + 1.10, respectively. These values were not significantly different from each other but were significantly lower (p < 0.02) than the value in late AR rabbits (Table 3). Tissue Distribution. Percentage injected dose per gram of tissue and relative concentrations in late AR, early AR, older control, and younger control rabbits are presented in Table 3. Because body weights and ages differed significantly between early AR and late AR, relative concentrations were derived by multiplying injected dose per gram by body weight. Consistent with the in vivo images, there was a strong trend toward significantly greater heart activity concentration in late AR than in older control rabbits (p = 0.057). No significant difference was found between early AR and younger control rabbits (p = 0.413). In addition, a significant correlation was observed between in vivo heart-to-lung ratio at 48 hours and ex vivo cardiac activity concentration in late AR and older control rabbits at death, immediately after the 48-hour imaging (in vivo heart-to-lung ratio vs ex vivo relative concentration, r = 0.81, p < 0.00001; in vivo heart-tolung ratio vs percentage injected dose per gram: r = 0.78, p = 0.0016). In contrast, no correlation was identified between in vivo heart-to-lung ratios and ex vivo relative concentration or percentage injected dose per gram in early AR and younger control rabbits.
DISCUSSION Our results indicate that external gamma camera imaging with radiolabeled antimyosin antibody can distinguish hearts with chronic severe hemodynamically compensated AR from those with subacute AR of similar severity and from normal hearts. In vivo images at 24 and 48 hours after isotope administration revealed consistently higher antimyosin antibody uptake in late AR rabbits than in other subgroups. These findings were supported by ex vivo demonstration of higher cardiac
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
29
24 l i t
Figure 2. Pinhole-collimator gamma camera images of thorax of late AR rabbit that manifested marked inferoapical antimyosin antibody uptake. Anterior images are shown at 24 hours after isotope administration (left) and 48 hours after isotope administration (right). Table 2. Results of anterior pinhole imaging: Count-per-pixel ratios 24 Hours Heart/lung Younger control (n = 8) Early AR (n = 11) Older control (n = 6) Late AR (n = 9)
1.81 2.04 2.06 2.70
_+ 0.47 ± 0.18 ± 0.15 ± 0.57*
48 Hours Heart/liver 0.76 0.78 0.77 0.99
+ 0.13 _+ 0.10 +_ 0.09 -+ 0.22*
Heart/lung 1.80 1.85 1.93 2.10
+ 0.17 _+ 0.17 _+ 0.13 ± 0.26*
Heart/liver 0.62 0.65 0.59 0.76
± 0.16 -+ 0.07 _+ 0.04 ± 0.19-1-
Values are m e a n + SD. *Significantly greater than corresponding older control, y o u n g e r control, and early AR values (p < 0.05). -]'Significantly greater than corresponding older control value (p < 0.05). There were no statistically significant differences at either 24 or 48 hours a m o n g early AR, y o u n g e r control, and older control values.
activity per unit cardiac mass, expressed as percentage injected dose per gram of heart, shortly after the 48-hour images were obtained. Previous in vitro studies have demonstrated that radiolabeled myosin-specific antibody binds intracellular myosin when cells have lost plasma membrane integrity? 6'37 Similar results have been obtained from in vivo experimental studies in animals in which radiolabeled antimyosin antibody has been found to associate with necrotic myocytes but not with normal myocytes. 14 Finally, supravital staining and conventional histologic methods have demonstrated myosin-specific antibody localization in areas of experimental myocardial infarction but not in the normal regions of the heart. ~3-15,38,39 Therefore, our finding suggests that myosin is exposed to the extracellular milieu in chronic severe experimental AR. Because cardiac myosin normally is present only
within cardiac myocytes, the left ventricular volume load chronically present in these animals appears to be associated with myocyte damage, which would be expected to accompany myocyte necrosis. This conclusion is consistent with our previous preliminary histologic observations identifying myocyte necrosis and associated fibrosis in many animals with chronic severe hemodynamically compensated experimental AR, and far more marked myocyte necrosis and fibrosis in animals in which congestive heart failure clearly is established. ~2 Our demonstration of the capacity of antimyosin imaging to detect myocyte damage in chronic experimental AR is consistent with clinical studies in which antimyosin imaging has been used in other settings. In a large clinical trial, antimyosin imaging has been highly accurate in identifying acute myocardial infarction] 6-t9 and both experimental and clinical studies have demon-
30
Lu et al. Noninvasive imaging in aortic regurgitation
Journal of Nuclear Cardiology January/February 1997, Part 1
Figure 3. Autoradiograms of 20 ~m-thick myocardial cryosections from younger control (shamoperated) rabbit (film exposure 11 days), older control (normal) rabbit (film exposure 10 days), early AR rabbit (film exposure 10 days), and late AR rabbit (film exposure 10 days). Each rabbit was killed 48 hours after administration of ~11in_labeled antimyosin antibody Fab fragment. In late AR rabbit, myocardial uptake and particularly subendocardial uptake are more intense than in other rabbits, consistent with in vivo gamma camera images. All images were obtained at same magnification and without contrast enhancement.
strated that the method accurately defines infarction site. ls-~9 Similarly, in comparison with analysis of myocardial biopsies, l alIn-labeled myosin-specific antibody uptake has accurately identified patients with acute myocarditis, 2°-23 a finding consistent with parallel studies in a murine myocarditis modelY ,26 Antimyosin antibody uptake also has been reported in patients with chronic idiopathic dilated cardiomyopathy and in those with hypertrophic cardiomyopathy, 26-3° in whom ongoing myocardial damage is evident. 26 Indeed, intensity of uptake in dilated cardiomyopathy carries prognostic information. 2s Although antimyosin antibody imaging has not been used clinically in AR, our findings are consistent with the reports of myocyte degeneration and fibrosis in surgical specimens from patients with severe chronic AR. a-8 Degenerated cells have demonstrated a wide variety of ultrastructural alterations, some of which would be expected to permit interaction between extracellular antibody and intracellular myosin. With respect to possible clinical application of antimyosin imaging in AR, it is important that the severity of myocyte degeneration
and fibrosis appears to be related to clinical and cardiac functional outcome. 1,5,7,1l Our previously published experience with the rabbit model used in these studies has indicated that it mirrors human disease relatively closely. Thus, regurgitant fraction and the resulting left ventricular dilatation and hypertrophy are similar in magnitude to those reported in severe AR in human beings. 32,33 In addition, as in human beings, some of the animals with AR predictably develop clinical heart failure and death within a definable period, and a subpopulation of those with heart failure manifest subnormal left ventricular performance at r e s t . 32'33 It is potentially important, therefore, that the rabbits in this study had not developed subnormal left ventricular performance or heart failure: this observation suggests that myosin exposure precedes these clinical and objective findings that in human beings presage untoward outcome. 21,28 Thus, radiolabeled antimyosin uptake may prove useful to predict the imminence of established prognostic indexes, with possible associated value in clinical decision making among patients with AR. The variability of antibody uptake seen in our animals is
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
31
Table 3. Heart activity of 11~ln-labeled antimyosin antibody (Myoscint) in AR and control rabbits Relative concentration (% ID/gm x body weight)
Heart activity (% ID/gm)
Younger control
0.2358 -- 0.1492 (n = 8)
0.0079 ___0.0050 (n = 8)
0.19 +- 0.05 O. 18 + 0.04 (n = 6)
1.05 _+ 0.10 (n = 6)
Early AR
0.3232 +_ 0.1355 (n = 11)
0.0105 _+ 0.0041 (n = 11)
0.30 + 0.06 0.27_+0.05(n= 11)
1.13 -+ 0.10 (n = 11)
Older control
0.1538 -+ 0.1132 (n = 6)
0.0042 _+ 0.0034 (n = 6)
O.15 -+ 0.05 0.14 + 0.04 (n = 5)
1.06 -+ 0.09 (n = 5)
Late AR
0.3583 -+ 0.2154" (n = 9)
0.0095 +- 0.0055* (n = 9)
0.23 -+ 0.07
1.27 +- 0.13t (n = 7)
Optical density /endocardium 1 \ epicardium /
0.18 -+ 0.05
(n = 7)
Optical density ratio (endocardium:epicardium)
ID, Injected dose. *p < 0.06 v s older control. "I"p < 0.02 vs older control, and y o u n g e r control, and early AR.
consistent with this inference, because it suggests that imaging results may mirror the observed variability of the pathophysiologic process and its outcome. Although these possibilities are intriguing, the limitations of our study preclude firm conclusions about potential clinical applicability. First, as yet we have no direct histopathologic data with which to compare our regional and global uptake findings. Thus, although our inferences as to the process underlying antimyosin antibody uptake is reasonable on the basis of previous studies, direct supporting evidence must be developed. Second, no animal had clinical congestive heart failure or subnormal left ventricular performance. Therefore it is not possible to associate the magnitude of antibody uptake directly with these conditions. Third, this crosssectional study did not assess the value of antibody uptake determination in predicting the subsequent development of heart failure or death. Finally, our data indicate a tendency toward reduction in the difference between cardiac uptake on 48-hour in vivo images in late AR versus control animals as compared with the difference on 24-hour images. Because blood pool activity, a major source of background, can be expected to decrease during the time between studies, and because ~11In is relatively long-lived, it might be expected that the uptake difference should have increased rather than decreased from the first study to the second. Lack of increase might raise concern about the reliability of external imaging for distinguishing AR-induced myocardial damage from normal myocardium. The latter concern cannot be adequately assessed without further study. Blood clearance of In-labeled antimyosin antibody was not determined
in our rabbit population. Therefore the effect of blood activity Clearance on heart-to-lung activity ratios at 24 and 48 hours is not clear. Given the potential variations of these pharmacokinetic factors, optimal imaging time cannot be determined intuitively and may be considerably earlier after isotope administration than would be expected with intact antibody imaging. Indeed, optimal imaging of myocardial infarction with radiolabeled monovalent Fab antimyosin antibody fragment can be achieved at 24 hours after administration, although blood pool interference at 24 hours may necessitate imaging at 48 hours in some patients. Irrespective of this unresolved issue, the difference between normal and late AR ratios in our study was statistically significant at both 24 and 48 hours, consistent with ex vivo cardiac uptake at 48 hours and supporting the principle that external imaging with radiolabeled antimyosin antibody i s potentially useful in detecting myocardial damage in experimental chronic AR. Because of the limitations noted above, this study can be regarded solely as generative of hypotheses. However, the clear association of antibody uptake with chronic severe AR before the development of heart failure or objective functional deterioration suggests the appropriateness of further study to assess the predictive value and pathophysiologic correlates of antimyosin antibody imaging in experimental AR and perhaps ultimately in the clinical setting. We thank Jeffrey Leppo, MD, for advice in the planning of this study; Bipim M. Mehta, PhD, for help in the pelformance of autoradiographic densitometry; and Ben Deyi, MPH, for statistical analysis.
32
Lu et al. Noninvasive imaging in aortic regurgitation
Journal of Nuclear Cardiology January/February 1997, Part 1
References 1. Maron B, Ferrans V, Roberts W. UltrastructuraI features of degenerated cardiac muscle cells in patients with cardiac hypertrophy. Am J Pathol 1975;79:387-413. 2. Maron B, Ferrans V, Roberts W. Myocardial ultrastmcmre in patients with chronic aortic valve disease. Am J Cardiol 1975;35:725-39. 3. Schwarz F, Flameng W, Schaper J, Henrlein F. Myocardial structure and function in patients with aortic valve disease and their relation to postoperative results. Am J Cardiol 1977;41:661-9. 4. Schwarz F, Flameng W, Schaper J, Henrlein F. Correlation between myocardial structure and diastolic properties of the heart in chronic aortic valve disease: effects of corrective surgery. Am J Cardiol 1978;42:895-903. 5. Fuster V, Danielson M, Robb R, Broadbent J, Brown A, Elveback L. Quantitation of left ventricular myocardial fiber hypertrophy and interstitial tissue in human hearts with chronically increased volume and pressure overload. Circulation 1977;55:504-8. 6. Hess O, Ritter M, Schneider J, Grimm J, Rurina M, Krayenbuebl H. Diastolic stiffness and myocardial structure in aortic valve disease before and after valve replacement. Circulation 1984;69: 855-65. 7. Krayenbuehl H, Hess O, Monrad E, Schneider J, Mall G, Rurina M. Left ventricular myocardial structure in aortic valve disease before, intermediate, and late after aortic valve replacement. Circulation 1988;79:744-55. 8. Villari B, Campbell SE, Hess OM, et al. Influence of collagen network on left ventricular systolic and diastolic function in aortic valve disease. J Am Coll Cardiol 1993;22:1477-84. 9. Weber K, Brilla C, Janicki J. Myocardial fibrosis: functional significance and regulatory factors. Cardiovasc Res 1993;27:341-8. 10. BriUa C, Janicki J, Weber K. Impaired diastolic function and coronary reserve in genetic hypertension. Circ Res 1991;69:107-15. 11. Villari B, Hess O, Kaufmann P, Krogmann O, Grimm J, Krayenbuehl H. Effect of aortic valve stenosis (pressure overload) and regurgitation (volume overload) on left ventricular systolic and diastolic function. Am J Cardiol 1992;69:927-34. 12. Liu SK, Magid NM, et al. Interstitial fibrosis and myocyte injury: relation to heart failure in aortic regurgitation [abstract]. Clin Res 1992;40:273A. 13. Khaw BA, Fallon JT, Strauss HW, Haber E. Myocardial infarct imaging of antibodies to canine cardiac myosin with indium-111diethylenetriamine pentaacetic acid. Science 1980;209:295-7. 14. Khaw BA, Fallon JT, Beller GA, Haber E. Specificity of the location of myosin specific antibody fragments in experimental myocardial infarction: histologic, histochemical, autoradiographic and scintigraphic studies. Circulation 1979;60:1527-31. 15. Khaw BA, Gold HK, Yasuda T, et al. Scintigraphic quantification of myocardial necrosis in patients after intravenous injection of myosin-specific antibody. Circulation 1986;74:501-8. 16. Berger H, Lahiri A, Leppo J, et al. Antimyosin imaging in patients with ischemic chest pain: Initial results of phase III multicenter trial [abstract]. J Nucl Med 1988;29:805-6. 17. Braat SH, deZwaan C, Teule J, Heidendal G, Wellens HJJ. Value of indium-ill monoclonal antimyosin antibody for imaging in acute myocardial infarction. Am J Cardiol 1987;60:725-6. 18. Johnson LL, Seldin DW, Becker LC, et al. Antimyosin imaging in acute transmural myocardial infarctions: results of a multicenter clinical trial. J Am Coll Cardiol 1989;13:27-35. 19. Volpini M, Giubbini R, Gei P, et al. Diagnosis of acute myocardial infarction by indium-111 antimyosin antibodies and correlation with the traditional techniques for the evaluation of extent and localization. Am J Cardiol 1989;63:7-13. 20. Yasuda T, Palacios IF, Dec W, et al. Indium 111-monoclonal
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36. 37.
38.
39.
antimyosin antibody imaging in the diagnosis of acute myocarditis. Circulation 1987;76:306-11. Dec GW, Palacios I, Yasuda T, et al. Antimyosin antibody cardiac imaging: its role in the diagnosis of myocarditis. J Am Coll Cardiol 1990;16:97-104. Carrio I, Bema L, Ballester M, et al. Indium-111 antimyosin scintigraphy to assess myocardial damage in patients with suspected myocarditis and cardiac rejection. J Nucl Med 1988;29:1893-900. Narula J, Khaw BA, Dec W Jr, et al. Recognition of acute myocarditis masquerading as acute myocardial infarction. N Engl J Med 1993;328:100-4. Yamada T, Matsumori A, Watanabe Y, et al. Pharmacokinetics of indium-l 1l-labeled antimyosin monoclonal antibody in murine experimental viral myocarditis. J Am Coll Cardiol 1990;16:1280-6. Kishimoto C, Hung GL, Ishibashi M, et al. Natural evolution of cardiac function, cardiac pathology and antimyosin scan in a murine myocarditis model. J Am Coll Cardiol 1991;17:821-7. Obrador D, Ballester M, Carrio I, Berna L, Pons-Llado G. High prevalence of myocardial monoclonal antimyosin antibody uptake in patients with chronic idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1989;13:1289-93. Obrador D, Ballester M, Carrio I, et al. Active myocardial damage without attending inflammatory response in idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1993;21:1667-71. Obrador D, Ballester M, Calrio I, et al. Presence, evolving changes, and prognostic implications of myocardial damage detected in idiopathic and alcoholic dilated cardiomyopathy by indium-Ill monoclonal antimyosin antibodies. Circulation 1994;89:2054-61. Nishimura T, Nagata S, Uehara T, Hayashida K, Mitani I, Kumita SI. Assessment of myocardial damage in dilated-phase hypertrophic cardiomyopathy by using indinm-111 antimyosin Fab myocardial scintigraphy. J Nucl Med 1991;32:1333-7. Nakata T, Gotoh M, Yonekura S, et al. Myocardial accumulation of monoclonal antimyosin Fab in hypertrophic cardiomyopathy and postpartum cardiomyopatby. J Nucl Med 1991;32:2291-4. Ballester-Rodes M, Carrio-Gasset I, Abadal-Befini L, Obrador-Mayol D, Bema-Roqueta L, Caralps-Riera JM. Patterns of evolution of myocyte damage after human heart transplantation detected by indium-111 monoclonai antimyosin. Am J Cardiol 1988;62:623-7. Magid NM, Opio G, Wallerson DC, Young MS, Borer JS. Heart failure due to chronic experimental aortic regurgitation. Am J Physiol 1994;267:H556-62. Magid NM, Wallerson DC, Borer JS, et al. Left ventricular diastolic and systolic performance during long-term follow-up of experimental aortic regurgitation. Am J Physiol 1992;263:H226-33. Magid NM, Young MS, Wallerson DC, et al. Hypertrophic and functional response to experimental chronic aortic regurgitation. J Mol Cell Cardiol 1988;20:239-46. Young MS, Magid NM, Wallerson DC, et al. Echocardiographic left ventricular mass measurement in small animals: anatomic validation in normal and aortic regurgitant rabbits. Am J Noninvasive Cardiol 1990;4:145-53. Kohler G, Milstein C. Continuous cultures of fused cells secreting antibody of predetermined specificity. Nature 1975;256:495-7. Khaw BA, Scott J, Fallon JT, Cahill SL, Haber E, Homcy C. Myocardial injury: quantitation by cell sorting initiated with antimyosin fluorescent spheres. Science 1982;217:1050-3. Khaw BA, Beller GA, Haber E, Smith TW. Localization of cardiac myosin-specific antibody in myocardial infarction. J Clin Invest 1976;58:439-46. Khaw BA, Gold HK, Leinbach RC, et al. Early imaging of experimental myocardial infarction by intracoronary administration of I-13 l-labeled anticardiac myosin (Fab') 2 fragments. Circulation 1978;58:1137-42.
From the Divisionof Cardiologyand the Divisionof NuclearMedicine, Cornell University Medical College, The New York HospitalCornell Medical Center, New York, N.Y. Supported in part by grant 1R29HL-40679-01, National Heart, Lung and Blood Institute (Bethesda, Md.); by grants from McNeil Pharmaceutical (Raritan, N.J.), Centocor (Malvern, Pa.) (which provided antimyosin antibody), The Howard Gilman Foundation (New York, N.Y.), The Daniel and Elaine Sargent CharitableTrust (New York, N.Y.), The David Margolis Foundation (New York, N.Y.), and The Charles and Jean Brunie Foundation (Bronxville, N.Y.); and by gifts from Rorlald and Jean Schiavone, Mary Jane Voute Arrigoni, and the late William Voute. During their participationin this study, Dr. Lu was a Howard Gilman FoundationAdvancedVisiting Fellow in CardiovascularResearch, Dr. Hardoff was a Howard Gilman Foundation Distinguished Visiting Scientist, and Dr. Gentile was a Howard Gilman Foundation Visiting Scientist at Cornell UniversityMedical College. Reprint requests: Jeffrey S. Borer, MD, The New York HospitalCornell Medical Center, 525 East 68th St., New York, NY 10021. Copyright © 1997 by American Society of Nuclear Cardiology. 1071-3581/97/$5.00+0 43/1/76175
Myocyte dropout and myocardial fibrosis have been observed in patients with aortic regurgitation (AR) in whom congestive heart failure develops and who undergo valve replacement. 1-8 These pathologic alterations are thought to contribute to the development of heart failure. 9-11 Similar histopathologic abnormalities have been reported in preliminary form in an experimental model of AR that closely simulates the characteristics and natural history of the human disease/2 Myocyte necrosis and fiber dropout and fibrosis were seen in most animals with chronic AR but were most prevalent in animals with congestive heart failure and were less marked or were absent among hemodynamically compensated animals. These observations suggest that in this model and perhaps in human beings, if fibrosis or myocyte necrosis could be identified noninvasively, the subsequent development of heart failure might be predictable. 25
26
Lu et al. Noninvasiveimaging in aortic regurgitation
Normally myosin is found entirely within the myocardial cell. Myocyte necrosis exposes myosin to the extracellular milieu. Therefore myocyte necrosis might be detected by external gamma camera imaging if radiolabeled antibody directed at myosin were administered. This technique has been used successfully to identify and localize myocyte necrosis due to myocardial infarction, myocarditis, and cardiomyopathy. 13-31 On the basis of these observations, we used antimyosin antibody imaging in animals with chronic severe experimental AR and in appropriate control animals to test the hypothesis that this noninvasive imaging technique can detect necrosis in myocardium subjected to prolonged volume loading in AR.
METHODS Animal Population. Thirty-four New Zealand White rabbits were studied. In 20 of the 34 rabbits, AR was surgically induced using closed-chest modification32 of our earlier procedures, 33'34 a s described briefly below. Of these 20, 9 (age 32.3 +_ 8.2 months, weight 3.80 +_ 0.30 kg) had moderate to severe AR (regurgitant fraction [RF] 56.6% + 7.2%) and were studied in the chronic phase, 98 to 128 weeks after AR induction. These rabbits did not show clinical signs of heart failure, and fractional shortening was 33.5 -+ 2.6%, within the limits we have previously defined as normal?3 The remaining 11 animals (age 6.4 _+ 0.9 months, weight 3.05 -+ 0.27 kg), with AR comparable in severity to that of the chronically regurgitant animals (RF 50.8 _+ 11.9% and fractional shortening 32.6 -+ 5.2%, no significant difference vs chronic AR), were studied 3 to 5 weeks after AR induction. This group was studied to enable accurate interpretation of results in animals with chronic AR by permitting determination of whether myocyte damage occurs early after AR development. Negative controls included normal uninstrumented animals and shamoperated rabbits prepared identically to AR animals except that the aortic valve was not perforated. Because there were significant differences in age and weight between the early and late AR groups, control animals were matched in age with each group. The younger control group (age 5.6 _+ 0.8 months, weight 2.95 + 0.27 kg) comprised 5 normal and 3 shamoperated rabbits; the older control group (age 33.7 -+ 8.9 months, weight 3.74 -+ 0.32 kg) included 3 normal and 3 sham-operated rabbits. Echocardiographic characteristics of AR and control rabbits are presented in Table 1. AR Induction. Rabbits were anesthetized with subcutaneous injection of 35 mg/kg ketamine and 5 mg/kg xylazine, with supplementary ketamine as needed during operation. The carotid artery was exposed and isolated. A 5F catheter was placed in the proximal carotid to monitor blood pressure and to act as a sheath for the introduction of a bevel-tipped catheter. The second catheter was advanced retrograde into the left ventricle through an aortic valve cusp during continuous arterial blood pressure monitoring. After the aortic valve was adequately perforated as determined by Doppler echocardiographic RF assessment (as discussed below) the catheter was
Journal of Nuclear Cardiology January/Febrnary1997, Part 1
withdrawn, the carotid ligated, and the incision closed in layers.31-35 Echocardiography. After induction of AR, aortic and pulmonary artery flows were measured by Doppler echocardiography to quantify RF, as previously described. 31-34 Twodimensional, M-mode, and Doppler echocardiograms were performed within 3 days after AR induction and once per month thereafter until within 1 week before antimyosin antibody imaging, according to methods previously described and validated. 3z-35Animals were sedated (35 mg/kg ketamine and 5 mg/kg xylazine); they were placed in left lateral decubitus; electrocardiographic leads were attached, and two-dimensionally directed M-mode echocardiograms of the left ventricle were obtained with a Doppler-echocardiographic system (Hewlett-Packard, Sonos 500 [Revision]) equipped with a 5-mHz, short-focus transducer and recorded on paper at 100 mm/sec. Fractional shortening was calculated, and left ventricular mass was determined as previously described and validated against autopsy measurements?T M Radionuclide Imaging. A Fab fragment of antimyosin monoclonal antibody coupled to diethylenetriaminepentaacetic acid (DTPA) (Myoscint, Centocor, Malvern, Pa.) was labeled with l llIn according to previously described methods. 13 Each rabbit received 500 Ixg of antibody labeled with 1 to 1.5 mCi l 1~in in 2.5 ml normal saline by bolus injection into an ear vein. Gamma camera images were acquired at 24 and 48 hours after antibody administration. Rabbits were sedated with intramuscular injection of 35 mg/kg ketamine and 5 mg/kg xylazine. With the rabbit placed and gently restrained in the supine position, chest images were obtained in the anterior and 45-degree left anterior oblique positions with a gamma camera (GE StarPort 400 AC) with 3.5 mm pinhole collimator. In addition, whole-body and fourfold-zoomed anterior and 45degree left anterior oblique images were acquired with a medium-energy parallel-hole collimator (GE StarCam 500A). In both cases, dual-photon acquisition of 300,000 to 500,000 counts was performed with photopeak energy windows of 173 keV -+ 15% and 247 keV -+ 15% and a 256 × 256 image matrix. To ensure comparability of technical factors associated with imaging, animals were paired or combined in small groups for isotope administration and imaging at the same session: in each pair or small group an AR animal was matched with at least one normal or sham-operated control animal. Image Analysis. Of the several images acquired, those obtained in the anterior position with the pinhole collimator exhibited the most pronounced myocardial "targeting" and heart-to-lung and heart-to-liver mean counts-per-pixel ratios. Therefore, these unprocessed images were used to quantitate antibody uptake and localization in the heart. A cardiac region of interest was defined and contained both ventricles. Regions of interest over the lung and liver were defined. Mean counts per pixel, were determined for each region, and mean heart-tolung and heart-to-liver counts-per-pixel ratios were calculated for the 24- and 48-hour images, as described by investigators. 2~,28,3° Interobserver count ratio variation was defined by two independent and blinded obseIvers. Intraobserver variation was calculated from consecutive determinations by a single
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
27
Table 1. E c h o c a r d i o g r a p h i c a n d a g e c h a r a c t e r i s t i c s o f A R a n d c o n t r o l rabbits at t i m e o f a n t i m y o s i n antibody imaging
Younger control
Early AR
Older control
Late AR
Age (mo)
Weight (kg)
l e f t ventricle internal diameter, diastole (cm)
Posterior wall thickness, diastole (cm)
Fractional shortening
5.6 _+ 0.8 (5.0-7.0) n=8 6.4 +_ 0.9 (5.0-7.8) n = II 33.7 _+ 8.9* (22.0-50.0) n=6 32.3 _+ 8.2"~ (20.0-50.0) n=9
2.95 -_ 0.27 (2.70-3.63) n=8 3.05 _+ 0.27 (2.66-3.60) n = II 3.74 _+ 0.32* (3.34-4.25) n=6 3.80 _+ 0 . 3 0 t (3.42-4.44) n=9
1.44 +_ 0.16 (1.20-1.67) n=7 1.88 +_ 0.34 (1.30-2.39) n = II 1.38 _+ 0.15 (1.20- 1.50) n=4 2.07 _+ 0.28:~ ( 1.60-2.50) n=9
0.23 _+ 0.06 (0.15-0.32) n=7 0.23 _+ 0.04 (0.18-0.31) n = II 0.26 + 0.03 (0.23-0.31 ) n=4 0.25 + 0.04 (0.20-0.34) n=9
32.50 _+ 5.10 (27.5-40.0) n=7 32.55 _+ 5.18 (23.1-37.5) n = II 36.00 _+ 7.07 (27.0-44.0) n=4 33.51 _+ 2.63 (31.6-39.0) n=9
(%)
*p < 0.02 vs younger control. t P < 0.001 vs early AIL ~p < 0.00l vs older control.
observer who was blinded to the identity of the rabbit being evaluated.
Autoradiography and Tissue Biodistribution. Rabbits were systemically heparinized (200 U/kg) and killed by intravenous injection of 100 mmol/L cadmium chloride immediately after the acquisition of the 48-hour images and while still anesthetized. In some experiments, catheters were placed in both the carotid artery and jugular vein so that the heart could be perfusion-fixed in situ, whereas in others the heart was immediately removed from the chest and perfused with a catheter placed directly into the root of the aorta. In either case, the heart was initially flushed with heparinized (10 U/ml) 100 mmol/L phosphate-buffered saline (pH 7.2) and subsequently fixed with either phosphate-buffered 4% paraformaldehyde or 2% glutaraldehyde and 2% paraformaldehyde. In general, no blood was apparent on the surface of the heart after either procedure. The fixed heart was cut into 3 to 4 mm slices from apex to base with a razor blade or a meat slicer. Alternating slices were used to determine radioactivity concentration in the heart and 1~~In antimyosin antibody distribution within the heart. Slices were placed in either 4% formalin or 15% sucrose-phosphatebuffered saline. Slices in 4% formalin were weighed and counted in an automated scintillation well counter calibrated for 111in" Net count rates were corrected for radioactive decay and converted to percentage of administered radioactivity per gram. Slices placed in sucrose were incubated at 4 ° C for -->3 hours, embedded in OTC medium (Tissue-Tek), frozen in isopentane at liquid nitrogen temperatures, and cut into 20 ixm-thick sections with a Haaker-Bright (Instrument Company, Huntingdon, England) cryostat to determine 111In antimyosin antibody localization autoradiographically. Approximately every tenth cryosection was exposed to scientific imaging film
(Kodak XO Mat AR) at - 7 0 ° C for 7 to 21 days. Intensifier screens were used to optimize images for analysis. As noted above, AR and control animals were matched for isotope administration and imaging; all members of these matched pairs or small groups were killed the same day and were placed on the same autoradiographic film to ensure identical exposure times for matched AR and control rabbits and thus for the four subgroups of AR and control animals. Autoradiograms were magnified and recorded with a high-resolution charge-coupled device camera (Sierra Scientific). The charge-coupled device image then was divided manually into endocardial and epicardial halves. The optical densities of the endocardial and epicardial regions were determined (MCID software version 4.2, Imaging Research). The ratios of endocardial-to-epicardial optical density were averaged for each rabbit, and the resulting values were used to calculate group averages for late AR, early AR, older control, and younger control rabbits. Statistical Analysis. Nonparametric Kruskal-Wallis one-way analysis of variance was used to compare heart-tolung and heart-to-liver counts-per-pixel ratios from in vivo images and to compare tissue concentrations of the heart and the endocardial-to-epicardial optical density ratios from autoradiograms. Late AR, early AR, older control, and younger control rabbits were compared.
RESULTS E c h o c a r d i o g r a p h i c f i n d i n g s . Table 1 presents the e c h o c a r d i o g r a p h i c findings in late AR, early A R , and older and y o u n g e r control animals. N o significant differ-
28
Lu et al. Noninvasive imaging in aortic regurgitation
48
Journal of Nuclear Cardiology January/February 1997, Part 1
I~r [
_ e, " t e.
qR
Figure 1. Anterior pinhole-collimator gamma camera images of thorax of younger control (normal), older control (shamoperated), early AR (acute AR), and late AR (chronic AR) rabbits 48 hours after injection of l~lIn-!abeled antimyosin antibody Fab fragment. At 48 hours, heart (center of each image) is better defined and reveals more intense isotope uptake relative to surrounding tissues (e.g., lungs and liver) in late AR rabbit than in early AR rabbit, younger control, or older control rabbit.
ences in fractional shortening were seen among the four groups. in Vivo Imaging. At 24 and at 48 hours after isotope administration, in vivo images revealed diffuse cardiac activity in all four groups (late AR, early AR, older control, and younger control) (Figure 1), except in one late AR rabbit that manifested marked inferoapical uptake (Figure 2). Heart-to-lung and heart-to-liver counts-per-pixel ratios were significantly greater in the late AR rabbits than in each of the three other groups at 24-hour imaging (Table 2). No significant differences were found among the three remaining groups, although younger control animals tended to exhibit a cardiac uptake lower than that of early AR rabbits. The heartto-lung ratios also were greater in late AR rabbits than in each of the three other groups at 48-hour imaging; for the heart-to-liver ratio, statistical significance was retained at 48 hours only for late AR versus older control animals (Table 2). Interobserver and intraobserver correlations were relatively strong for in vivo gamma camera measures of the heart, lung, and liver uptake: correlation coefficients were 0.997, 0.794, and 0.972 (all p < 0.001), respectively, between two blinded independent observers, and
were 0.98, 0.88, and 0.83 (allp < 0.01), respectively, for two consecutive determinations by a single observer. Autoradiographic Distribution. The imaging results were mirrored in the autoradiographic findings. Intense uptake was seen only in late AR animals (Figure 3). Moreover, radiotracer uptake was most prominent in the subendocardium of animals with late AR; such regional sequestration was not seen in normal or shamoperated animals of either age group or in early AR animals. In late AR rabbits, the average endocardial-toepicardial uptake ratio was 1.27 _+ 0.13 vs 1.06 + 0.09 in older controls (p < 0.02). Late AR rabbits also manifested significantly greater absolute subendocardial uptake than did age-matched control rabbits (p = 0.02) (Table 3). Ratios in early AR and younger control animals were 1.13 _+ 0.10 and 1.05 + 1.10, respectively. These values were not significantly different from each other but were significantly lower (p < 0.02) than the value in late AR rabbits (Table 3). Tissue Distribution. Percentage injected dose per gram of tissue and relative concentrations in late AR, early AR, older control, and younger control rabbits are presented in Table 3. Because body weights and ages differed significantly between early AR and late AR, relative concentrations were derived by multiplying injected dose per gram by body weight. Consistent with the in vivo images, there was a strong trend toward significantly greater heart activity concentration in late AR than in older control rabbits (p = 0.057). No significant difference was found between early AR and younger control rabbits (p = 0.413). In addition, a significant correlation was observed between in vivo heart-to-lung ratio at 48 hours and ex vivo cardiac activity concentration in late AR and older control rabbits at death, immediately after the 48-hour imaging (in vivo heart-to-lung ratio vs ex vivo relative concentration, r = 0.81, p < 0.00001; in vivo heart-tolung ratio vs percentage injected dose per gram: r = 0.78, p = 0.0016). In contrast, no correlation was identified between in vivo heart-to-lung ratios and ex vivo relative concentration or percentage injected dose per gram in early AR and younger control rabbits.
DISCUSSION Our results indicate that external gamma camera imaging with radiolabeled antimyosin antibody can distinguish hearts with chronic severe hemodynamically compensated AR from those with subacute AR of similar severity and from normal hearts. In vivo images at 24 and 48 hours after isotope administration revealed consistently higher antimyosin antibody uptake in late AR rabbits than in other subgroups. These findings were supported by ex vivo demonstration of higher cardiac
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
29
24 l i t
Figure 2. Pinhole-collimator gamma camera images of thorax of late AR rabbit that manifested marked inferoapical antimyosin antibody uptake. Anterior images are shown at 24 hours after isotope administration (left) and 48 hours after isotope administration (right). Table 2. Results of anterior pinhole imaging: Count-per-pixel ratios 24 Hours Heart/lung Younger control (n = 8) Early AR (n = 11) Older control (n = 6) Late AR (n = 9)
1.81 2.04 2.06 2.70
_+ 0.47 ± 0.18 ± 0.15 ± 0.57*
48 Hours Heart/liver 0.76 0.78 0.77 0.99
+ 0.13 _+ 0.10 +_ 0.09 -+ 0.22*
Heart/lung 1.80 1.85 1.93 2.10
+ 0.17 _+ 0.17 _+ 0.13 ± 0.26*
Heart/liver 0.62 0.65 0.59 0.76
± 0.16 -+ 0.07 _+ 0.04 ± 0.19-1-
Values are m e a n + SD. *Significantly greater than corresponding older control, y o u n g e r control, and early AR values (p < 0.05). -]'Significantly greater than corresponding older control value (p < 0.05). There were no statistically significant differences at either 24 or 48 hours a m o n g early AR, y o u n g e r control, and older control values.
activity per unit cardiac mass, expressed as percentage injected dose per gram of heart, shortly after the 48-hour images were obtained. Previous in vitro studies have demonstrated that radiolabeled myosin-specific antibody binds intracellular myosin when cells have lost plasma membrane integrity? 6'37 Similar results have been obtained from in vivo experimental studies in animals in which radiolabeled antimyosin antibody has been found to associate with necrotic myocytes but not with normal myocytes. 14 Finally, supravital staining and conventional histologic methods have demonstrated myosin-specific antibody localization in areas of experimental myocardial infarction but not in the normal regions of the heart. ~3-15,38,39 Therefore, our finding suggests that myosin is exposed to the extracellular milieu in chronic severe experimental AR. Because cardiac myosin normally is present only
within cardiac myocytes, the left ventricular volume load chronically present in these animals appears to be associated with myocyte damage, which would be expected to accompany myocyte necrosis. This conclusion is consistent with our previous preliminary histologic observations identifying myocyte necrosis and associated fibrosis in many animals with chronic severe hemodynamically compensated experimental AR, and far more marked myocyte necrosis and fibrosis in animals in which congestive heart failure clearly is established. ~2 Our demonstration of the capacity of antimyosin imaging to detect myocyte damage in chronic experimental AR is consistent with clinical studies in which antimyosin imaging has been used in other settings. In a large clinical trial, antimyosin imaging has been highly accurate in identifying acute myocardial infarction] 6-t9 and both experimental and clinical studies have demon-
30
Lu et al. Noninvasive imaging in aortic regurgitation
Journal of Nuclear Cardiology January/February 1997, Part 1
Figure 3. Autoradiograms of 20 ~m-thick myocardial cryosections from younger control (shamoperated) rabbit (film exposure 11 days), older control (normal) rabbit (film exposure 10 days), early AR rabbit (film exposure 10 days), and late AR rabbit (film exposure 10 days). Each rabbit was killed 48 hours after administration of ~11in_labeled antimyosin antibody Fab fragment. In late AR rabbit, myocardial uptake and particularly subendocardial uptake are more intense than in other rabbits, consistent with in vivo gamma camera images. All images were obtained at same magnification and without contrast enhancement.
strated that the method accurately defines infarction site. ls-~9 Similarly, in comparison with analysis of myocardial biopsies, l alIn-labeled myosin-specific antibody uptake has accurately identified patients with acute myocarditis, 2°-23 a finding consistent with parallel studies in a murine myocarditis modelY ,26 Antimyosin antibody uptake also has been reported in patients with chronic idiopathic dilated cardiomyopathy and in those with hypertrophic cardiomyopathy, 26-3° in whom ongoing myocardial damage is evident. 26 Indeed, intensity of uptake in dilated cardiomyopathy carries prognostic information. 2s Although antimyosin antibody imaging has not been used clinically in AR, our findings are consistent with the reports of myocyte degeneration and fibrosis in surgical specimens from patients with severe chronic AR. a-8 Degenerated cells have demonstrated a wide variety of ultrastructural alterations, some of which would be expected to permit interaction between extracellular antibody and intracellular myosin. With respect to possible clinical application of antimyosin imaging in AR, it is important that the severity of myocyte degeneration
and fibrosis appears to be related to clinical and cardiac functional outcome. 1,5,7,1l Our previously published experience with the rabbit model used in these studies has indicated that it mirrors human disease relatively closely. Thus, regurgitant fraction and the resulting left ventricular dilatation and hypertrophy are similar in magnitude to those reported in severe AR in human beings. 32,33 In addition, as in human beings, some of the animals with AR predictably develop clinical heart failure and death within a definable period, and a subpopulation of those with heart failure manifest subnormal left ventricular performance at r e s t . 32'33 It is potentially important, therefore, that the rabbits in this study had not developed subnormal left ventricular performance or heart failure: this observation suggests that myosin exposure precedes these clinical and objective findings that in human beings presage untoward outcome. 21,28 Thus, radiolabeled antimyosin uptake may prove useful to predict the imminence of established prognostic indexes, with possible associated value in clinical decision making among patients with AR. The variability of antibody uptake seen in our animals is
Journal of Nuclear Cardiology Volume 4, Number 1;25-32
Lu et al. Noninvasive imaging in aortic regurgitation
31
Table 3. Heart activity of 11~ln-labeled antimyosin antibody (Myoscint) in AR and control rabbits Relative concentration (% ID/gm x body weight)
Heart activity (% ID/gm)
Younger control
0.2358 -- 0.1492 (n = 8)
0.0079 ___0.0050 (n = 8)
0.19 +- 0.05 O. 18 + 0.04 (n = 6)
1.05 _+ 0.10 (n = 6)
Early AR
0.3232 +_ 0.1355 (n = 11)
0.0105 _+ 0.0041 (n = 11)
0.30 + 0.06 0.27_+0.05(n= 11)
1.13 -+ 0.10 (n = 11)
Older control
0.1538 -+ 0.1132 (n = 6)
0.0042 _+ 0.0034 (n = 6)
O.15 -+ 0.05 0.14 + 0.04 (n = 5)
1.06 -+ 0.09 (n = 5)
Late AR
0.3583 -+ 0.2154" (n = 9)
0.0095 +- 0.0055* (n = 9)
0.23 -+ 0.07
1.27 +- 0.13t (n = 7)
Optical density /endocardium 1 \ epicardium /
0.18 -+ 0.05
(n = 7)
Optical density ratio (endocardium:epicardium)
ID, Injected dose. *p < 0.06 v s older control. "I"p < 0.02 vs older control, and y o u n g e r control, and early AR.
consistent with this inference, because it suggests that imaging results may mirror the observed variability of the pathophysiologic process and its outcome. Although these possibilities are intriguing, the limitations of our study preclude firm conclusions about potential clinical applicability. First, as yet we have no direct histopathologic data with which to compare our regional and global uptake findings. Thus, although our inferences as to the process underlying antimyosin antibody uptake is reasonable on the basis of previous studies, direct supporting evidence must be developed. Second, no animal had clinical congestive heart failure or subnormal left ventricular performance. Therefore it is not possible to associate the magnitude of antibody uptake directly with these conditions. Third, this crosssectional study did not assess the value of antibody uptake determination in predicting the subsequent development of heart failure or death. Finally, our data indicate a tendency toward reduction in the difference between cardiac uptake on 48-hour in vivo images in late AR versus control animals as compared with the difference on 24-hour images. Because blood pool activity, a major source of background, can be expected to decrease during the time between studies, and because ~11In is relatively long-lived, it might be expected that the uptake difference should have increased rather than decreased from the first study to the second. Lack of increase might raise concern about the reliability of external imaging for distinguishing AR-induced myocardial damage from normal myocardium. The latter concern cannot be adequately assessed without further study. Blood clearance of In-labeled antimyosin antibody was not determined
in our rabbit population. Therefore the effect of blood activity Clearance on heart-to-lung activity ratios at 24 and 48 hours is not clear. Given the potential variations of these pharmacokinetic factors, optimal imaging time cannot be determined intuitively and may be considerably earlier after isotope administration than would be expected with intact antibody imaging. Indeed, optimal imaging of myocardial infarction with radiolabeled monovalent Fab antimyosin antibody fragment can be achieved at 24 hours after administration, although blood pool interference at 24 hours may necessitate imaging at 48 hours in some patients. Irrespective of this unresolved issue, the difference between normal and late AR ratios in our study was statistically significant at both 24 and 48 hours, consistent with ex vivo cardiac uptake at 48 hours and supporting the principle that external imaging with radiolabeled antimyosin antibody i s potentially useful in detecting myocardial damage in experimental chronic AR. Because of the limitations noted above, this study can be regarded solely as generative of hypotheses. However, the clear association of antibody uptake with chronic severe AR before the development of heart failure or objective functional deterioration suggests the appropriateness of further study to assess the predictive value and pathophysiologic correlates of antimyosin antibody imaging in experimental AR and perhaps ultimately in the clinical setting. We thank Jeffrey Leppo, MD, for advice in the planning of this study; Bipim M. Mehta, PhD, for help in the pelformance of autoradiographic densitometry; and Ben Deyi, MPH, for statistical analysis.
32
Lu et al. Noninvasive imaging in aortic regurgitation
Journal of Nuclear Cardiology January/February 1997, Part 1
References 1. Maron B, Ferrans V, Roberts W. UltrastructuraI features of degenerated cardiac muscle cells in patients with cardiac hypertrophy. Am J Pathol 1975;79:387-413. 2. Maron B, Ferrans V, Roberts W. Myocardial ultrastmcmre in patients with chronic aortic valve disease. Am J Cardiol 1975;35:725-39. 3. Schwarz F, Flameng W, Schaper J, Henrlein F. Myocardial structure and function in patients with aortic valve disease and their relation to postoperative results. Am J Cardiol 1977;41:661-9. 4. Schwarz F, Flameng W, Schaper J, Henrlein F. Correlation between myocardial structure and diastolic properties of the heart in chronic aortic valve disease: effects of corrective surgery. Am J Cardiol 1978;42:895-903. 5. Fuster V, Danielson M, Robb R, Broadbent J, Brown A, Elveback L. Quantitation of left ventricular myocardial fiber hypertrophy and interstitial tissue in human hearts with chronically increased volume and pressure overload. Circulation 1977;55:504-8. 6. Hess O, Ritter M, Schneider J, Grimm J, Rurina M, Krayenbuebl H. Diastolic stiffness and myocardial structure in aortic valve disease before and after valve replacement. Circulation 1984;69: 855-65. 7. Krayenbuehl H, Hess O, Monrad E, Schneider J, Mall G, Rurina M. Left ventricular myocardial structure in aortic valve disease before, intermediate, and late after aortic valve replacement. Circulation 1988;79:744-55. 8. Villari B, Campbell SE, Hess OM, et al. Influence of collagen network on left ventricular systolic and diastolic function in aortic valve disease. J Am Coll Cardiol 1993;22:1477-84. 9. Weber K, Brilla C, Janicki J. Myocardial fibrosis: functional significance and regulatory factors. Cardiovasc Res 1993;27:341-8. 10. BriUa C, Janicki J, Weber K. Impaired diastolic function and coronary reserve in genetic hypertension. Circ Res 1991;69:107-15. 11. Villari B, Hess O, Kaufmann P, Krogmann O, Grimm J, Krayenbuehl H. Effect of aortic valve stenosis (pressure overload) and regurgitation (volume overload) on left ventricular systolic and diastolic function. Am J Cardiol 1992;69:927-34. 12. Liu SK, Magid NM, et al. Interstitial fibrosis and myocyte injury: relation to heart failure in aortic regurgitation [abstract]. Clin Res 1992;40:273A. 13. Khaw BA, Fallon JT, Strauss HW, Haber E. Myocardial infarct imaging of antibodies to canine cardiac myosin with indium-111diethylenetriamine pentaacetic acid. Science 1980;209:295-7. 14. Khaw BA, Fallon JT, Beller GA, Haber E. Specificity of the location of myosin specific antibody fragments in experimental myocardial infarction: histologic, histochemical, autoradiographic and scintigraphic studies. Circulation 1979;60:1527-31. 15. Khaw BA, Gold HK, Yasuda T, et al. Scintigraphic quantification of myocardial necrosis in patients after intravenous injection of myosin-specific antibody. Circulation 1986;74:501-8. 16. Berger H, Lahiri A, Leppo J, et al. Antimyosin imaging in patients with ischemic chest pain: Initial results of phase III multicenter trial [abstract]. J Nucl Med 1988;29:805-6. 17. Braat SH, deZwaan C, Teule J, Heidendal G, Wellens HJJ. Value of indium-ill monoclonal antimyosin antibody for imaging in acute myocardial infarction. Am J Cardiol 1987;60:725-6. 18. Johnson LL, Seldin DW, Becker LC, et al. Antimyosin imaging in acute transmural myocardial infarctions: results of a multicenter clinical trial. J Am Coll Cardiol 1989;13:27-35. 19. Volpini M, Giubbini R, Gei P, et al. Diagnosis of acute myocardial infarction by indium-111 antimyosin antibodies and correlation with the traditional techniques for the evaluation of extent and localization. Am J Cardiol 1989;63:7-13. 20. Yasuda T, Palacios IF, Dec W, et al. Indium 111-monoclonal
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36. 37.
38.
39.
antimyosin antibody imaging in the diagnosis of acute myocarditis. Circulation 1987;76:306-11. Dec GW, Palacios I, Yasuda T, et al. Antimyosin antibody cardiac imaging: its role in the diagnosis of myocarditis. J Am Coll Cardiol 1990;16:97-104. Carrio I, Bema L, Ballester M, et al. Indium-111 antimyosin scintigraphy to assess myocardial damage in patients with suspected myocarditis and cardiac rejection. J Nucl Med 1988;29:1893-900. Narula J, Khaw BA, Dec W Jr, et al. Recognition of acute myocarditis masquerading as acute myocardial infarction. N Engl J Med 1993;328:100-4. Yamada T, Matsumori A, Watanabe Y, et al. Pharmacokinetics of indium-l 1l-labeled antimyosin monoclonal antibody in murine experimental viral myocarditis. J Am Coll Cardiol 1990;16:1280-6. Kishimoto C, Hung GL, Ishibashi M, et al. Natural evolution of cardiac function, cardiac pathology and antimyosin scan in a murine myocarditis model. J Am Coll Cardiol 1991;17:821-7. Obrador D, Ballester M, Carrio I, Berna L, Pons-Llado G. High prevalence of myocardial monoclonal antimyosin antibody uptake in patients with chronic idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1989;13:1289-93. Obrador D, Ballester M, Carrio I, et al. Active myocardial damage without attending inflammatory response in idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1993;21:1667-71. Obrador D, Ballester M, Calrio I, et al. Presence, evolving changes, and prognostic implications of myocardial damage detected in idiopathic and alcoholic dilated cardiomyopathy by indium-Ill monoclonal antimyosin antibodies. Circulation 1994;89:2054-61. Nishimura T, Nagata S, Uehara T, Hayashida K, Mitani I, Kumita SI. Assessment of myocardial damage in dilated-phase hypertrophic cardiomyopathy by using indinm-111 antimyosin Fab myocardial scintigraphy. J Nucl Med 1991;32:1333-7. Nakata T, Gotoh M, Yonekura S, et al. Myocardial accumulation of monoclonal antimyosin Fab in hypertrophic cardiomyopathy and postpartum cardiomyopatby. J Nucl Med 1991;32:2291-4. Ballester-Rodes M, Carrio-Gasset I, Abadal-Befini L, Obrador-Mayol D, Bema-Roqueta L, Caralps-Riera JM. Patterns of evolution of myocyte damage after human heart transplantation detected by indium-111 monoclonai antimyosin. Am J Cardiol 1988;62:623-7. Magid NM, Opio G, Wallerson DC, Young MS, Borer JS. Heart failure due to chronic experimental aortic regurgitation. Am J Physiol 1994;267:H556-62. Magid NM, Wallerson DC, Borer JS, et al. Left ventricular diastolic and systolic performance during long-term follow-up of experimental aortic regurgitation. Am J Physiol 1992;263:H226-33. Magid NM, Young MS, Wallerson DC, et al. Hypertrophic and functional response to experimental chronic aortic regurgitation. J Mol Cell Cardiol 1988;20:239-46. Young MS, Magid NM, Wallerson DC, et al. Echocardiographic left ventricular mass measurement in small animals: anatomic validation in normal and aortic regurgitant rabbits. Am J Noninvasive Cardiol 1990;4:145-53. Kohler G, Milstein C. Continuous cultures of fused cells secreting antibody of predetermined specificity. Nature 1975;256:495-7. Khaw BA, Scott J, Fallon JT, Cahill SL, Haber E, Homcy C. Myocardial injury: quantitation by cell sorting initiated with antimyosin fluorescent spheres. Science 1982;217:1050-3. Khaw BA, Beller GA, Haber E, Smith TW. Localization of cardiac myosin-specific antibody in myocardial infarction. J Clin Invest 1976;58:439-46. Khaw BA, Gold HK, Leinbach RC, et al. Early imaging of experimental myocardial infarction by intracoronary administration of I-13 l-labeled anticardiac myosin (Fab') 2 fragments. Circulation 1978;58:1137-42.