ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 10, No.3, 307-314, 1996 
Comparative study of 201Tl-scintigraphic image and myocardial pathologic findings in patients with dilated cardiomyopathy 
Lin-xue LL* Ryuji NOHARA,* Kazumi OKUDA,** Ryohei HOSOKAWA,* Tatsuhiko HATA,* Masaru TANAKA,* Akira MATSUMORI,* Masatoshi FUJITA,*** Nagara TAMAKI,**** Junji KONISHI**** and Shigetake SASAYAMA* 
* Third Division, Department of Internal Medicine, Kyoto University, ** Takeda Hospital, Kyoto, ***College of Medical Technology, Kyoto University, ****Department of Nuclear Medicine, Kyoto University 
The objective of the present study was to characterize the production of 201Tl myocardial perfusion defects, the relation between the 201Tl multiple small defects and the myocardial damage indicated by myocardial fibrosis shown histopathologically in patients with dilated cardiomyopathy (DCM). 
Rest 201Tl scintigraphy was performed in thirty-seven patients with myocardial tissue fibrosis by endomyocardial biopsy, and without stenosis of the coronary artery. 201Tl myocardial SPECT images were visually classified into 4 grades according to the severity of inhomogeneous perfusion defects (IPD), O: none, l: slight, 2: moderate, 3: severe. 201Tl uptake, defect regions (DR), and coefficient of variation % (CV%) were also quantified by Bull's eye quantification in nineteen patients. During cardiac catheterization, three biopsy specimens were obtained from the lateral wall to the apical region of the left ventricle and the amount of fibrosis was assessed by means of light microscopic morphometry. The myocardial fibrosis was also classified into 4 grades by a point-counting method. Autopsy study was also assessed in six patients. 201Tl perfusion defects were observed in 35 (94.6%) patients, of whom 29 (78.4%) showed inhomogeneous perfusion defects. Twenty-four (64.9,%) showed Stage O and 1 201Tl findings, and 21 (62.2%) had myocardial fibrosis in stage 1 . Clinically, the correlation between the grades of the IPD, % 201Tl uptake, DR and CV% of myocardial uptake, which were calculated semiquantitatively by Bull's eye image, and the histological grades of fibrosis were also good (IPD vs. fibrosis: r = 0.7014; % 201Tl uptake vs. fibrosis: r = -0.6542; DR vs. fibrosis: r = 0.7027; CV% vs. fibrosis: r = 0.6985). The 201Tl SPECT findings were in close agreement with the severity of myocardial fibrosis confirmed by autopsy, but the grading of the IPD was not related to the ejection fraction or left ventricular diameter. 
It showed a higher rate of inhomogeneous 201Tl myocardial perfusion defects (78.4%) in patients with DCM. This result may contribute to the clinical evaluation of DCM or differentiation from other diseases. Furthermore, the grading of 201Tl inhomogeneous perfusion defects related to the myocardial fibrosis of left ventricular myocardium may contribute to speculation of the myocardial degenerative stage in clinical settings. 
Key words : dilated cardiomyopathy (DCM), 201Tl-SPECT, myocardial fibrosis. 
INTRODUCTION 
THALLIUM-201 myocardial single photon emission computed tomography (201Tl SPECT) is often reported to show multiple perfusion defects in patients with dilated cardiomyopathy (DCM) which are different from the defects observed in coronary artery disease.1,2 There have also been studies on the methods to express the defect patterns and the degrees of abnormal uptake on 201Tl myocardial SPECT imaging.3 The autopsy studies with DCM showed that the percentage of fibrosis in the left ventricular wall was significantly greater than that with secondary eccentric hypertrophy or normal heart,4,5 and there was significant relationship between % fibrosis and 201Tl defect in the non-uniform pattern high fibrosis group, and the areas with a high incidence of % fibrosis showed (201)Tl defects.(6) But there is no study on the relationships of histopathologic fibrosis of biopsy to the degrees and patterns of SPECT imaging. Sufficient data on the correspondence between (201)Tl imaging and histopathological findings are needed for the solution of this problem. The purposes of the present study are 1) to describe the patterns of thallium scintigraphy in patients with dilated cardiomyopathy, and 2) to correlate the scintigraphic findings with endomyocardial biopsy findings. 
METHODS 
Patients: 
The study group comprised 37 patients (25 male and 12 female, mean age 49 +- 16 years old) with DCM. The subjects were the patients admitted to the Third Department of Internal Medicine, Kyoto University Hospital, from 1987 to 1995, and were diagnosed as having DCM on the basis of the history, chest X-rays, ECG, echocardiography and left ventriculography. The diagnosis of DCM was suspected when patients had an increased cardiothoracic ratio and unexplained congestive heart failure. Echocardiography and catheterization data supported DCM, when they showed diffuse hypokinesis in wall motion and normal coronary arteries without valvular heart disease. The final decision was made when biopsy findings showed cellular loss and fibrosis.(7) Patients with complications such as arrhythmia were included. Patients underwent (201)Tl SPECT within 9 +- 5 weeks before left ventricular biopsies without intercurrent changes in their clinical status. The left ventricular ejection fraction was determined by the Kennedy's(8) method by left ventriculography obtained by cardiac catheterization. The left ventricular end-diastolic dimension was determined in the left ventricular long-axial view of standard two-dimensional echocardiography at the level of the papillary muscles (Table 1). 
(201)Tl imaging: 
Resting (201)Tl myocardial SPECT was performed for all the patients after a 12 hour fast. After 15 minutes of intravenous (201)Tl injection with 74-111 MBq (2-3 mCi), SPECT scans were obtained with a gamma camera equipped with a General Electric 400 AC/T linked to a MaxiStar computer before 1990. After 1990, SPECT imaging was performed with General Electric Medical Systems Limited Albans Heart England, linked to a YNO 189 YOKOKAWA Medical System. Collimator used was a low-energy and general purpose type. Data were collected from 32 views of 30 sec each over 180deg from the 45deg right anterior oblique angle to the 45deg left posterior oblique angle. Oblique tomograms parallel to the long-and short-axis of the left ventricle were also reconstructed. For(9,10) nineteen of them the Bull's eye map quantification (15 males and 4 females, 50 +- 16 years old) was serially performed. In the other patients, the Bull's eye map quantification could not be performed because the computer system was different. 
Quantitative image analysis: 
For each study, the left ventricular myocardium was divided into nine segments on the SPECT images and polar map display (basal anterior, septal, inferior, lateral, middle anterior, septal, inferior, lateral and apical) to assess the (201)Tl score (Fig.1). The perfusion defect of (201)Tl tracer in each segment was scored by the consensus of the three experienced observers using a four-point grading system, 0: normal, 1: slight inhomogeneous perfusion abnormality (multiple small hypoperfusion),2:moderate inhomogeneous defect (multiple small defect within the size of 1 segment. The stage is between 1 and 3), and 3: severe inhomogeneous defect or larger defect (multiple large defect greater than 1 segment or a quite large solitary defect)(3,9) without having any information about the patients (Fig.2). For quantitative analysis, a circumferential profile curve was generated from apical to basal short-axial slices to a Bull's eye polar map with 100% as the maximum count in each (201)Tl image. Defect regions (DR), that were (201)Tl uptake < 80% in the nine myocardial segments were set manually on the (201)Tl image. Defect regions were defined as those with a total size of (201)Tl uptake < 80% in the nine myocardial segments. Within each region of interest, the mean % (201)Tl uptake of the tracer which was (201)Tl uptake < 80% in nine myocardial segments were calculated. The coefficient of variance % (CV%) was calculated as standard deviation / mean x 100.(10-15) 
Left ventricular biopsies: 
Myocardial biopsy was performed from the left ventricular endocardium during cardiac catheterization in all subjects. Three biopsy specimens were obtained from the area of the left ventricle, from the@lateral wall to the apical region. Myocardial slices were stained by Masson's Trichrome. Morphometric evaluations were performed by two investigators, who were unaware of the patients scintigraphic data, and the amount of fibrosis was assessed by standard light microscopy. The ratio of the fibrotic area to the total tissue area was calculated as the percent area of fibrosis (%F), and the severity of left venincular fibrosis (LVF) was graded histopathologically from Stage 0 to Stage 3 (0 = trivial; %F from 0 to 25%, 1 = mild %F from 26 to 50%, 2 = moderate; %F from 51 to 75%, 3 = severe; %F > 75%) (Fig. 3). The severity of endocardial fibrosis was studied by autopsy in 6 patients who died of DCM (3 males and 3 females, mean age 61 +- 5 years). 
STATISTICAL ANALYSIS 
Data are given as the mean +- SD. Comparison between the regression lines was performed by student's t test. The correlation between the (201)Tl score and left ventricular fibrosis (LVF) was determined from linear regression analysis. Correlations with the % (201)Tl uptake vs. LVF, DR vs. LVF, CV% vs. LVF, (201)Tl score vs. LVEF and (201)Tl score vs. LVDd were calculated by the (nonparametric) Spearman's rank statistics. For all tests, a value p < 0.05 was considered indicative of a statistically significant difference. 
RESULTS 
Relation between (201)Tl SPECT imaging and myocardial biopsy: 
Of the 37 patients, (201)Tl perfusion defects were observed in 35 (94.6%) patients, of whom 29 (78.4%) had inhomogeneous defects, 2 patients (5.4%) had normal perfusion and 6 (16.2%) had larger defects. Stages 0 and 1 were the most frequent in 24 (64.8%) patients. 
Myocardial fibrosis was demonstrated by histological examination of biopsy specimens in 34 (100%), and 23 (62.2%) of the patients in stage 1. Stage 2 and stage 3 were less frequent (37.8%). The correlation between the grad-ing of 201Tl SPECT findings and the biopsy staging of fibrosis was good (r = 0.7014, p < 0.001) in DCM (Fig. 4). When patients were studied by the quantification of Bull's eye images, the correlation between fibrosis and % 201Tl uptake, DR and CV% were all good (LVF vs. % 201Tl uptake, r = -0.6542; LVF vs. DR, r = 0.7027; LVF vs. CV%, r = 0.6985) (Fig. 5). 

Relationship of 201Tl SPECT findings to LVEF and LVDd: The 201Tl myocardial SPECT stage of inhomogeneity showed no correlation with LVEF (r = - 0.3241 ) or LVDd (r = 0.2517). 
Comparison between autopsyfindings and 201Tl SPECT findings:
Concerning the 6 patients who died of DCM, Intervals between autopsy and 201Tl SPECT were 17 to 31 months (mean = 21.1 +- 6 months), and the grading of the 201Tl perfusion defect was related to autopsy findings (Table 3). 

Case presentation 
CASE 1 (autopsy No. 1. Fig. 6-a). A 43 year-old-male patient with cardiomyopathy died 16 months after 201Tl SPECT imaging. The heart weighed 420 g, and the left ventricle was dilated and hypertrophied. 201Tl SPECT imaging showed noticeable left ventricular enlargement, severe inhomogeneous defect and large perfusion defects in the anterior, septal and inferior segments. Autopsy revealed remarkable massive fibrosis in the same region of the left ventricle. 
CASE 2 (autopsy No. 2. Fig. 6-b). A 70-year-old-male with cardiomyopathy died 20 months after 201Tl SPECT imaging. The heart weighed 520 g, all four chambers were dilated and showed signs of diffuse interstitial fibrosis. 201Tl SPECT finding in this patient showed dilatation of the left ventricle and moderate inhomogeneous defects. 

DISCUSSION 

Our study indicated that an inhomogeneous 201Tl-SPECT image will show at least myocardial fibrotic changes depending on the stage of the DCM, by means of a comparative study with endomyocardial biopsy. List of abbreviations: 
201Tl SPECT: 201Tl single photon emission computed tomography; DCM: dilated cardiomyopathy; IPD: inhomogeneous perfusion defect; LVF: left ventricular fibrosis; %F: % fibrosis; SD: standard deviation; DR: defect regions; CV%: coefficient of variation%; LVEF: left ventricular ejection fraction; LVDd: Ieft ventricular diastolic dimension; ECG: electrocardiogram There have been previous reports assessing patients with DCM by means of 201Tl scintigraphy.1-4 Doi et al.4 classified dipyridamole myocardial 201Tl scintigraphy findings into 4 stages, that is, no defect, small solitary defects, multiple small defects, and large defects, showing that 44.2% of the patients had multiple small defects. In the present study, multiple defects were observed in 78.4% of the patients, and the defect pattern was completely different from that observed in myocardial infarction. The causative mechanism underlying the production of 201Tl perfusion defects in patients with DCM remains obscure with the manifestation of myocardial fibrosis. There have been many theories as to 201Tl defects such as myocardial degeneration, atrophy, necrosis and ventricular wall thinning. In this study, about 60% of patients were found to have a slight 201Tl defect and slight fibrosis, and it is possible that some myocardial damage might have been assessed by 201Tl findings. We had conflicting results regarding the effect of 201Tl finding to fibrosis. 1) Two patients had slight fibrosis but normal 201Tl perfusion. It may be for this reason that a lesser degree of fibrosis cannot respond to the 201Tl uptake. 2) Two patients had severe fixed 201Tl defects, but lesser degrees of fibrosis. A cause and effect relationship cannot be proved by myocardial fibrosis. It is probable that the production of fixed 201Tl defects by other factors such as ventricular wall thinning caused by myocardial atrophy or metabolical change ma exert fixed 201Tl defects. 
There have been several reports on endomyocardial biopsies and autopsied hearts in patients with DCM.4,6,16 Yaoita6 compared histopathological findings in cardiac muscle and myocardial perfusion defects on 201Tl scintigraphy in ten autopsied cases with DCM. They showed a significant relationship between % fibrosis and 201Tl defects in a high % fibrosis group, and the areas with a high incidence of % fibrosis showed 201Tl defects. But Dunn et al.1 compared the 201Tl score with autopsy findings in DCM patients and found that the ventricular wall was thin and no fibrosis was observed at the site of 201Tl defects. In our study, however, the severity of fibrosis of endocardial tissues was in close agreement with the stage of defects in the 201Tl myocardial scintigrams. We classified 201Tl defects into 4 stages, and examined the relationships of perfusion defects and defect patterns of 201Tl myocardial SPECT with left ventricular histopathologic fibrosis. The classification of 201Tl showed a good correlation with fibrotic change in the left ventricle. 201Tl perfusion defect patterns have therefore been strongly related with advancement of myocardial fibrosis in patients with DCM. Autopsy findings also support the relationship between the 201Tl score and myocardial fibrosis. 
The redistribution of 201Tl scintigraphy in patients with DCM is hardly detectable.1-3 Dunn et al.1 performed exercise 201Tl scanning, in which 40% of patients with cardiomyopathy and 67% with coronary artery disease showed redistribution on a 4-hour scan. In patients with DCM, reversible defects could not be explained by regional differences in coronary blood flow. Thallium uptake by myocardial cells also depends on factors other than coronary blood flow. In cardiomyopathy, the myocardial cell membrane in some areas of the heart may be abnormal, resulting in inhomogeneous uptake of thallium. Reversible defects may also result from the change in left ventricular geometry between the time of the exercise and the redistribution scans. Perfusion defect on thallium imaging can occur in patients with idiopathic dilated cardiomyopathy and chronic heart failure, but it cannot be a definitive index in patients with chronic heart failure to distinguish coronary artery disease from cardiomyopathy even if a complete defect is present. In our study, reversible defects were not evaluated, because the majority of the patients could not perform the exercise test and the number of subjects was insufficient. 
It is difficult to express the area of abnormal uptake with our scoring method, and this method may not necessarily detect the magnitude of abnormal uptake. The interpretation of 201Tl myocardial SPECT images is a subjective process, and whether it corresponds with the degree of histological fibrosis or not is open to question. In this study, the Bull's eye method has also been used to evaluate defects semiquantitatively in order to obtain a clear answer. These findings showed a high correlation between % 201Tl uptake, DR, CV% and biopsy of fibrosis with DCM. Thus visual findings were supported by a semiquantitative method in our study. Because the data available in this study were filed over the long period, not all of the studies were possible to evaluate semiquantitatively by the Bull's eye method, but serial studies that were done by the Bull's eye method provided clear results as shown here. 
DCM is a disease characterized by enlargement of the left ventricular cavity and reduction in myocardial con-tractility. In many DCM patients, the wall motion was diffusely reduced, resulting in a characteristic decrease in left ventricular function,17-19 unlike old myocardial infarction (OMI), which shows signs of synchrony. No significant correlation was reported to be found between 201Tl myocardial uptake and left ventricular ejection fraction (LVEF) in patients with DCM.20 In our study, no correlation was also observed between the defects of resting 201Tl SPECT and LVEF or LVDd. Our data indicate that dilation of the left ventricle may lead to apical thinning, and apical defects of 201Tl imaging, but no patchy perfusion defect was detected in patients with aortic valve insufficiency showing similar decreases in LVEF and LVDd with DCM. Our classification by means of 201Tl images is therefore useful for evaluation and differentiation between DCM and aortic valve insufficiently with low LVEF and/or a large left ventricular cavity. STUDY LIMITATIONS 

There have been some limitations to this study. The first limitation is that biopsy specimens are taken in a very restricted area of the heart. It is clinically impossible to take samples of the myocardium from all around the ventricle as in a 201Tl SPECT image. For this reason, direct comparison between 201Tl findings and those of biopsies may be difficult, but Onodera et al. reported that the percent area of fibrosis in the septum, anterior, lateral and posterior walls of the left ventricle in DCM showed no significant differences among groups.4This suggests that biopsy findings may represent to some extent total left ventricular fibrosis, showing that comparison of biopsy specimens with the 201Tl image is reasonable. Secondly, this study only assessed the correlation between 201Tl defects and myocardial fibrosis. The possibility of the production of perfusion defects by other factors in patients with DCM, such as metabolical change, should be further evaluated. Thirdly, in the present study, the redistribution image was not assessed, because the number of subjects was insufficient. Redistribution, perhaps related to thallium extraction and viable myocardial muscle, may exist at rest or be induced by exercise, and will produce perfusion defects in the 201Tl image. The possibility of a different image in the redistribution phase will be less, but further evaluation is necessary to understand the effect of redistribution. Lastly, this retrospective study was performed over a long period, and it is possible that differences in assessing images because of differences in the types of devices used and in image quality. 

CONCLUSIONS 

Perfusion defects were observed by 201Tl myocardial SPECT in 94.6% of the patients with DCM in whom fibrosis was demonstrated by biopsy. 201Tl myocardial SPECT showed inhomogeneous perfusion defects in many patients (78.4%). The appearance of multiple small defects in myocardial SPECT was in close agreement with histological detection of myocardial fibrosis in DCM. Evaluation by the Bull's eye semi-quantitative method supported these findings. The grading of perfusion defects in 201Tl myocardial SPECT imaging was also in close agreement with the autopsy findings in endomyocardial fibrosis. The inhomogeneous perfusion defects on the 201Tl SPECT imaging are therefore considered to be one of the markers of myocardial fibrosis, which is not totally dependent on cardiac function, although further evaluation is needed for more reasonable grading. These results may contribute to the clinical evaluation of DCM or differentiation from other disease. The above grading of perfusion defects on 201Tl myocardial SPECT imaging may be useful in determining histopathological changes, including fibrosis in patients with DCM. 

REFERENCES 

1 . Dunn RF, Uren RF, Sadick N, Bautovich G, Mclaughlin A, Hiroe M. Comparison of thallium-201 scanning in idiopathic dilated cardiomyopathy and severe coronary artery disease. Circulation 66: 804-8 10, 1982. 
2. Saltissie S, Hockings B, Croft DN, Webb-peploe MM. Thallium-201 myocardial imaging in patients with dilated and ischemic cardiomyopathy. Br Heart J 46: 290-295, 1981. 
3. Doi Y, Chikamori T, Takata J, Yonezawa Y, Poloniecki JD, Ozawa T, et al. Prognostic value of thallium-201 perfusion defects in idiopathic dilated cardiomyopathy. Am J Cardiol 67: 188-193, 1991. 
4. Onodera T, Fujiwara H, Tanaka M, Wu D-J, Hamashina Y, Kawai C, et al. Quantitative analysis of cardiac fibrosis in normal hearts, hearts with secondary eccentric hypertrophy and hearts with dilated cardiomyopathy. Card-Ang Bull (Cardiologisch-Angiologisches Bulletin) 3: 53-59, 1986. 
5. Schwarz F. Mall G, Zebe H, Blick J. Derks H, Berks H, et al. Quantitative morphologic findings of the myocardium in idiopathic dilated cardiomyopathy. Am J Cardiol 51 : 501-506, 1983. 
6. Yaoita H. A clinicopathological study on dilated cardiomyopathy. Fukushima lgakuzashi 38 (1): 113-121 , 1988. (in Japanese) 
7. Report of the WHO/ISFC task: force on the definition and classification of cardiomyopathies. Br Heart J 44: 672-673, 1980. 
8. Kennedy JW, Trenholme SE, Kasser IS. Left ventricular volume and mass from single-plane cineangiocardiograms: A comparison of antero-posterior and right anterior oblique methods. Am Heart J 80: 343-351 , 1970. 
9. Tamaki N, Ohtani H, Yonekura Y, Nohara R, Kambara H, Kawai C, et al. Significance of fill-in after thallium-201 reinjection following delayed imaging: comparison with regional wall motion and angiographic findings. J Nucl Med 31: 1617-1623, 1990. 
10. Tamaki N, Yonekura Y. Mukai T, Kodama S, Kadota K, Kambara H, et al. Stress thallium-201 transaxial emission computed tomography; quantitative versus qualitative analysis for evaluation of coronary artery disease. J Am Coll Cardiol 4: 1213-1221 , 1984. 
11. Tamaki N, Ohtani H, Yamashita K, Magata Y, Yonekura Y, Nohara R, et al. Metabolic activity in the areas of new fill-in after thallium-201 reinjection: comparison with positron emission tomography using Fluorine-18-Deoxyglucose. J Nucl Med 32: 673-678, 1991 . 
12. Tamaki S. Nakajima H, Murakami T, Yui Y, Kambara H, Kadota K, et al. Estimation of infarct size by myocardial emission computed tomography with thallium-201 and its relation to creatine kinase-MB release after myocardial infarction in man. Circulation 66: 994-1001 , 1982. 
l 3. Eisner R, Churchwell A, Noever T, Nowak D, Cloninger K, Dunn D, et al. Quantitative analysis of the tomography thallium-201 myocardial Bullseye display: critical role of correcting for patient motion. J Nucl Med 29: 91-97, 1988. 
14. Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull's eye images.J Nucl Med 33: 272-276,1992. 
15. Zimmermann R, Mall G, Rauch B, Zimmer G, Gabel M, Zehelein J, et al. Residual (201)Tl activity in irreversible defects as a marker of myocardial viability clinicopathological study. Circulation 91 : 1016-1021, 1995. 
16. Fujiwara H, Onodera T, Tanaka M, Shirane H, Kato H, Yoshikawa J, et al. Progression from hypertrophic obstructive cardiomyopathy to typical dilated cardiomyopathylike features in the end stage. Jpn Circ J 48: 1210=1214, 1984. 
17. Nakayama Y, Shimizu G, Hirota Y, Saito T, Kino M, Kitaura Y, et al. Functional and histopathologic correlation in patients with dilated cardiomyopathy: an integrated evaluation by multivariate analysis.J Am Coll Cardiol 10:186-192,1987.18. Hayashida W, Kumada T, Nohara R, Tanio H, Kambayashi M, Ishikawa N, et al. Left ventricular regional wall stress in dilated cardiomyopath . Circulation 82: 2075-2083, 1990. 
19. Sunnerhagen KS, Bhargava V, Shabetai R. Regional left ventricular wall motion abnormalities in idiopathic dilated cardiomyopath . Am J Cardiol 65: 361-370, 1990. 
20. Yamakado K, Takeda K, Kitano T, Nakagawa T, Futagami Y, Konishi J, et al. Serial change of iodine-123 metaiodobenzylguanidine (MIBG) myocardial concentration in patients with dilated cardiomyopathy. Eur J Nucl Med 19: 265-270, 1992.