J. Comp. Path. 2004, Vol.

130, 235245

www.elsevier.com/locate/jcpa

Morphometry of Bovine Dilated Cardiomyopathy

P. Nart, A. Williams, H. Thompson and G. T. Innocent*
Department of Veterinary Pathology, *Comparative Epidemiology and Informatics Group, Institute of Comparative Medicine, Glasgow University Veterinary School, Bearsden Road, Glasgow G61 1QH, UK

Summary
Bovine dilated cardiomyopathy (BDCM) is a primary disease of the myocardium that has been described in Holstein-Friesian cattle worldwide in the last 20 years. The principal morphological changes in the myocardium are interstitial brosis and increased variability in cardiomyocyte size. Sections of heart muscle from nine cases of BDCM and nine unaffected controls matched for age, sex and breed were studied by means of a computer-assisted image analyser to measure the degree of brosis, and the cardiomyocyte cellular and nuclear cross-sectional area and length. The amount of connective tissue in the hearts of BDCM cases was increased by 6.7 times, the nuclear transverse cross-sectional area by 1.9 times, and the cardiomyocyte length and cross-sectional area by 1.7 and 1.6 times, respectively. This resulted in an estimated 2.5-fold increase in mean cardiomyocyte volume. Animals with clinical signs of BDCM showed a mean loss of 51% of the total number of cardiomyocytes as compared with controls. Of the ve parameters studied, the percentage of brosis was found to be the most consistent discriminator for BDCM. It is possible that the degree of brosis could be used to distinguish BDCM from other cardiac diseases of cattle. q 2003 Elsevier Ltd. All rights reserved.
Keywords: cardiomyocyte morphometry; cardiomyopathy; cattle; myocardium

Introduction
The rst published report of bovine dilated cardiomyopathy (BDCM) in Holstein-Friesian cattle came from Japan (Sonoda et al., 1982). It was also reported in Switzerland (Martig et al., 1982), Canada (where cases had been recorded since 1971; Baird et al., 1986), Sweden (Olsson, 1987), Australia (McLennan and Kelly, 1990), the United Kingdom (Bradley et al., 1991) and Denmark (Leifsson et al., 1994). The main clinical signs of BDCM are those associated with severe congestive heart failure, and the gross pathological ndings include a dilated heart, congested liver, ventral subcutaneous oedema and ascites or hydrothorax. Histologically, the heart shows diffuse brosis, and the cardiomyocytes focal degeneration,
Correspondence to: A. Williams, Department of Pathology and Infectious Diseases, Royal Veterinary College, Hawkshead Campus, Hawkshead Lane, North Mymms, Hertfordshire AL9 7TA, UK. 00219975/$ - see front matter doi: 10.1016/j.jcpa.2003.11.002

extensive vacuolation and variability in calibre (Bradley et al., 1991). The aetiology of BDCM in the Holstein-Friesian breed is still unknown, but Dolf et al. (1998) proposed that an autosomal recessive gene might play a role. Dilated cardiomyopathy (DCM) in man may also follow a familial pattern (Towbin and Bowles, 2002) and BDCM has been proposed as a model of human DCM because of similarities in clinical features and pattern of protein expression (Eschenhagen et al., 1995; Weekes et al., 1999). Morphometry is the process of producing, from images of two-dimensional histological sections, measurements from which three-dimensional volume, surface area, and number and length of tissue components can be determined. The tissues functional capacity can then be estimated by quantifying the volume fraction of the parenchyma and stroma and their proportional relationships (Loud and Anversa, 1984). Morphometric studies
q 2003 Elsevier Ltd. All rights reserved.

236

P. Nart et al.

can be undertaken with a light or an electron microscope and the data obtained either by computer-assisted image analysis or by the more traditional point-counting method with a square grid. In the latter, intersections of the grid (points) falling on features of interest are counted and expressed either as a percentage or as a proportion of total points. However, computerassisted image analysis measures larger areas in a shorter time (Beltrami et al., 1996) and the results agree closely with those obtained by the classic point counting method (Porzio et al., 1995); it is therefore often the method of choice. Morphometric analysis of human DCM has demonstrated increased brosis, cardiomyocyte hypertrophy and bre elongation. Estimates of the degree of brosis in DCM in man range from 14 to 20%, as compared with a baseline of 3% in control hearts (Dick et al., 1982; Unverferth et al., 1986; Beltrami et al., 1995; Ohtani et al., 1995). Studies of cardiomyocyte length in DCM showed increases of 40% (Gerdes, 1992) and 59% (Beltrami et al., 1995) in man and of 35% (Spinale et al., 1991) and 50% (Kajstura et al., 1995) in experimental models of pacemaker-induced DCM in dogs. Cardiomyocyte hypertrophy, which may be measured as increases in width, diameter or crosssectional area, has been reported in human DCM as 50% (Rowan et al., 1988), 30% (Dick et al., 1982; Unverferth et al., 1986) and 20% (Beltrami et al., 1995). Doubling of the nuclear cross-sectional area was reported by Rowan et al. (1988), and an increase in cardiomyocyte nuclear size was related by Figulla et al. (1985) and Pelliccia et al. (1994) to decreased functional status and poor prognosis. Left ventricular failure was associated with an estimated 39% loss of ventricular cardiomyocytes (Kajstura et al., 1995), and with a positive correlation between the extent of cardiomyocyte loss and cellular hypertrophy in samples taken at necropsy. However, there appeared to be little if any association between histological changes seen on cardiac biopsy in human DCM and clinical signs and prognosis (Baandrup et al., 1981); this may reect the fact that biopsies are usually taken from the right ventricle, whereas cardiac function is more dependent on the left ventricle. Nevertheless, the degree of brosis is correlated with a reduction in the ejection fraction or contractibility of the heart (Schwarz et al., 1983; Ohtani et al., 1995). Cardiomyocyte hypertrophy and interstitial brosis are the main qualitative morphological ndings in both human and bovine DCM (Robinson and Ferrans, 1975; Furuoka et al., 2001). However, no previous studies have quantied

cardiomyocyte hypertrophy and interstitial brosis in BDCM. The purpose of the present study, therefore, was to determine the main quantitative morphological changes in BDCM by objective statistical analysis and to identify the most consistent and relevant histological ndings of this cardiac disorder. The total numbers of cardiomyocytes in the hearts of animals with BDCM and control animals were estimated from the measurements in an attempt to understand the pathological mechanisms of this disease.

Materials and Methods

Tissues Examined Samples from the outer (free) wall of the left ventricle of the heart were collected from nine Holstein-Friesian cattle (nos 1 9) in which BDCM had been diagnosed at the Glasgow University Veterinary School Pathology Department over a period of 9 years. The diagnosis was made on the basis of clinical ndings and gross pathology, followed by histopathological conrmation. Control material was obtained from apparently normal hearts of nine unaffected animals (nos 10 18) matched for age, sex and breed with the diseased animals slaughtered at an abattoir. To estimate the degree of brosis in control and affected hearts, longitudinal sections of myocardium stained with Sirius red and picric acid (Junqueira et al., 1979) were examined by light microscopy, and a bit map image was developed from the original image by means of the KS 300 V.3 software package (Image Associates/Zeiss; Oberkochen, Germany). The percentage of brosis was calculated by comparing the proportion of brous tissue (Sirius red positive, Fig. 1a; recognized as white in the bit map image) with the total of all tissue (black and white) (Fig. 1b). Estimations of cardiomyocyte area, nuclear area and nuclear length were made from sections stained by haematoxylin and eosin (HE) (Bancroft and Stevens, 1996). Areas with transverse sections of myobres were selected. The contour of the bres was then drawn manually (as shown in Fig. 1c). The same method was used for nuclear area and nuclear length. Estimates of cardiomyocyte length were made on sections stained with phosphotungstic acid haematoxylin (PTAH), which clearly demonstrates intercalated discs (Bancroft and Stevens, 1996) (Fig. 1d). Quantitative measurements of area and length were taken automatically by a multipurpose colour image processor with the KS 300 V.3 software package. Two recorded

Bovine Dilated Cardiomyopathy

237

Fig. 1a d. (a) Sirius red staining of myocardium of case 1 showing brous tissue stained red. (b) Computer-generated bit map of Fig. 1a rendered black and white to enable the computer to calculate the percentage of brosis (now white). (c) HEstained section of the myobres of case 1 in transverse section, and the hand drawn contours to delineate cross-sectional area. (d) PTAH staining of cardiomyocytes in longitudinal section with hand-drawn line between intercalated discs. Bar, 200 mm (a,b); 20 mm (c); 100 mm (d).

command sequences (macros) were created for analysing data. Numbers of Measurements Made and Statistical Analysis A pilot study was rst undertaken to determine the number of observations required to distinguish between affected and control animals in respect of the various cardiomyocyte parameters. Six samples taken from each of four healthy animals and all nine cases of BDCM were measured for degree of brosis and cardiomyocyte cross-sectional area. Application of Bartletts test (Paradine and Rivett, 1960) indicated that although the values for unaffected animals were homoscedastic (equal variance), the values for the animals with BDCM

had variances that were different from the healthy animals and from each other. Standard tests for differences between groups require that all groups have the same variance. Due to the difculties in calculating power when the variances between and within groups vary, the variances in the pilot study were used to calculate minimum signicant differences in a t-test with Welchs correction for heteroscedasticity. The numbers of observations were chosen with a signicance level of P , 0:05 in mind. Thus, 30 measurements of cardiomyocyte length, nuclear length and nuclear area and ve measurements of brosis per animal were judged to be sufcient to distinguish between groups. Crosssectional areas of 60 cardiomyocytes per animal

238

P. Nart et al.

were measured, however, because of the greater variability of this parameter. Histological sections from all 18 animals were then sampled systematically, since systematic sampling yields smaller errors than random sampling in histometrics (Ebbeson and Tang, 1967). In the present study, constraints arising from the random orientation of the cardiomyocytes within the section were overcome by careful selection of appropriate elds (i.e., those in which the myocytes were aligned transversely or longitudinally to the section plane), thereby ensuring accurate systematic measurements. Since the pilot study had identied the heteroscedastic nature of the data, and since repeated measures were taken for each animal, a method of analysis was required that was suitable for such data. One such method is that of iterative generalized least squares, a generalization of the least squares method for normally distributed data, which provides maximum likelihood estimators for multilevel mixed-effects models (Goldstein, 1986, 1989; Ihaka and Gentleman, 1996). This is implemented in the gls function in the R statistical package that was used to analyse the data arising from the morphometrical analyses in this study (Ihaka and

Gentleman, 1996). In the statistical tests, signicance was set at the 5% level. The analyses of measurements of each parameter for each animal were then converted into a boxplot representation (see Figs 2 6). In these boxplots, the extremes of the box indicate the 25th and 75th quartiles. Values that were more than twice (vertical lines, ending with an open horizontal bar) the interquartile range away from the median were considered outliers and marked with an open circle. The whiskers at each end of the plots extend to the extremes of the non-outlier data. The traverse line across the boxplot represents the median value. Statistical analysis was used to determine if there were signicant differences between diseased and control animals for the ve parameters studied. Animals with BDCM that appeared to differ less markedly from the controls than the rest of the affected group were then tested to determine if they were indeed signicantly different from the controls. The sub-groups derived from comparing these animals against the control group are marked with a lled triangle or lled circle that indicates the value of the mean for that particular sub-group. The heart weight was recorded in four BDCM cases and four controls. The heart weights were

Fig. 2. Percentage of brosis. Vertical axis: brosis as percentage of the total area examined in histological sections. Horizontal axis: bovine dilated cardiomyopathy (BDCM) cases (animals 1 9); controls (animals 10 18). A subgroup composed of animals 6 and 9 (triangles) was also compared with the control group. The horizontal dotted lines represent the weighted mean of the animals with BDCM and of control animals.

Bovine Dilated Cardiomyopathy

239

Fig. 3. Cardiomyocyte cross-sectional area. Vertical axis: area expressed in square microns. Horizontal axis: bovine dilated cardiomyopathy (BDCM) cases (animals 1 9); controls (animals 10 18). The selected subgroup for comparison with the controls was composed of cases 3 and 6 (triangles). The horizontal dotted lines represent the weighted mean of BDCM and control groups.

compared between groups by means of a Mann Witney U -test. The total volume of the myocardium (tissue volume) was determined by dividing its weight by the specic gravity of muscle tissue, 1.06 g/ml (Mendez and Keys, 1960; Loud and Anversa, 1984). An approximation of cardiomyocyte volume was then made, assuming a cylindrical form, from the mean values of area and length for the BDCM and control groups. The total number of cardiomyocytes contained in the heart tissue was then calculated by the formula:

taken from the hearts of each animal for the ve parameters studied are shown in Figs 2 6. Fibrosis There was signicantly P , 0:01 more brosis in the nine animals with BDCM mean 22% than in the nine controls mean 3% (Fig. 2). As animals 6 and 9 had values that were clearly lower than those of the other animals with BDCM, a second analysis was conducted. A statistical difference was also found when animals 6 and 9 (as one subgroup) were compared with the controls P , 0:01: Cardiomyocyte Cross-sectional Area A signicant difference P , 0:01 was found between the cardiomyocyte cross-sectional area for the group of nine animals with BDCM (mean 599 mm2) and the group of nine control animals (mean 366 mm2) (Fig. 3). However, considerable variation was seen between individual animals with BDCM, and there was a clear overlap in values between some animals in the BDCM

Total numbers of cardiomyocytes Heart volume1 2 percentage of fibrosis=100 : Mean cardiomyocyte volume

Results
The values obtained for amount of interbrillar collagen, and length and cross sectional area of the cardiomyocyte and its nucleus are summarized in Table 1 and boxplots of the values of measurements

240

P. Nart et al.

Fig. 4. Cardiomyocyte length. Vertical axis: the length of cardiomyocytes expressed in microns. Horizontal axis: bovine dilated cardiomyopathy (BDCM) cases (animals 1 9); controls (animals 10 18). The subgroup of BDCM cases compared with the controls consisted of case 7 (triangle). The horizontal dotted lines represent the weighted mean of BDCM and control groups.

group and control animals. Thus, comparison of a subgroup composed of animals 3 and 6 with the group of nine controls showed no signicant difference P 0:32: Cardiomyocyte Length The values for the nine animals with BDCM (mean 139 mm) and the nine control animals (mean 86 mm), differed signicantly P , 0:01 (Fig. 4). As animal 7 had lower values than did the other eight animals with BDCM, the values for this one animal were compared to the group of nine controls; the difference was still statistically signicant P , 0:01: Cardiomyocyte Nuclear Area The cardiomyocyte nuclear cross-sectional area in cattle with BDCM (mean 26 mm2) differed signicantly P , 0:01 from that in the control group (mean 14 mm 2) (Fig. 5). Inspection of the boxplots showed that the nine affected animals could be placed in one of three subgroups. Animals 1, 3, 4 and 6 (forming one subgroup) were then compared with the controls, as was the subgroup

formed by animals 7, 8 and 9. In both instances, a statistically signicant difference was found between the subgroup of BDCM cases and the controls P , 0:01: Nuclear Length No signicant differences P 0:15 were found between animals with BDCM and the controls (Fig. 6), the mean nuclear lengths for which were 14 mm and 13 mm, respectively. Overlapping of Measurements Data ranking enabled the percentage of measurements in the BDCM group that exceeded the highest value obtained for the control group to be calculated. These results are presented in Table 2. Measurements of Cardiomyocyte Volume An approximation of cardiomyocyte volume was calculated, assuming a cylindrical form, from the mean values for area and length for the BDCM and control groups. The mean values obtained for the cardiomyoctye cell volume in BDCM cases and

Bovine Dilated Cardiomyopathy

241

Fig. 5. Cardiomyocyte nuclear area. Vertical axis: area expressed in microns. Horizontal axis: bovine dilated cardiomyopathy (BDCM) cases (animals 1 9); controls (animals 10 18). Subgroups selected for comparison with the controls were: subgroup one (cases 1,3,4 and 6) (triangles); subgroup two (cases 7, 8 and 9) (circles). The horizontal dotted lines represent the weighted mean of BDCM and control groups.

controls were 139 087 mm3 and 54 700 mm 3, respectively. Measurements of Cardiomyocyte Numbers The heart weights of four BDCM and four control animals did not differ signicantly P 0:67: Similarly, no signicant differences were found in terms of heart tissue volume. The estimated total number of cardiomyocytes is given in Table 3; the mean loss for the four cases of BDCM was calculated to be 51%.

Discussion
This study represents the rst attempt made to quantify and analyse statistically the main histological characteristics of BDCM. There were signicant increases in percentage brosis, in cardiomyocyte length and cross-sectional area, and in cardiomyocyte nuclear cross-sectional area. A mean loss of 51% cardiomyocytes was calculated in animals with BDCM compared with unaffected animals matched for age, sex and breed. The main microscopical characteristics of BDCM were ranked by their value

in distinguishing affected from control animals; the percentage area of brosis gave the best discrimination, followed by cardiomyocyte length, cardiomyocyte area and nuclear area. These changes can be considered as the main morphometric features of the terminal stage of BDCM; as brosis was the only characteristic that did not present any overlapping of values, it is proposed as the most reliable discriminatory parameter for BDCM. Further studies are clearly required to investigate to what extent the characteristics analysed in this work are altered in other cardiac diseases of cattle; it is considered, however, that the degree of brosis observed in BDCM is much greater than that present in other bovine cardiac diseases such as cor pulmonale and congenital heart disease. Although the affected and control groups differed signicantly in respect of cardiomyocyte size, there was some overlap in the values for individual animals (Fig. 3). Thus, cardiomyocyte hypertrophy does not appear to be a reliable diagnostic parameter for BDCM. Similarly, although elongation of nuclei is considered a sign of hypertrophy, there was no evidence of

242

P. Nart et al.

Fig. 6. Nuclear length. Vertical axis: length expressed in microns. Horizontal axis: bovine dilated cardiomyopathy (BDCM) cases (animals 1 9); controls (animals 10 18).

an increase in the length of cardiomyocyte nuclei in BDCM cases in this study. In cardiac disease causing a reduction in cardiac output, the heart maintains systemic perfusion by utilising two related mechanisms (Frank-Starling principle). Thus there is an increase in resting sarcomere and muscle length to allow a greater preload reserve, and this muscle elongation produces an increase in muscle contractibility that results in increased stroke volume (Strobeck and Sonnenblick, 1986). These two adaptative mechanisms are exacerbated in DCM, but the elongation of individual cardiomyocytes does not necessarily imply overall tissue hypertrophy as DCM is caused by a degenerative process. The extreme hypertrophy of the cardiomyocytes in some BDCM cases

compared to others may indicate that such animals have had a longer clinical illness, allowing more time for compensatory changes to occur (Benjamin et al., 1981). However, insufcient clinical information was available to test this hypothesis in the present study. Moreover, the small numbers of cases prevented conrmation of any possible relation between cardiomyocyte loss and hypertrophy in BDCM, as described in human DCM (Kajstura et al., 1995). Genetic studies indicate that BDCM has an autosomal recessive mode of inheritance in combination with some unidentied environmental factor (Baird et al., 1986; Satoh, 1988; Bradley et al., 1991; Dolf et al., 1998). It would be of interest to determine if animals heterozygous for BDCM but

Table 1 Median, mean and standard deviation values for the ve parameters studied in the BDCM and control groups Control group Parameter Fibrosis (%) Cardiomyocyte area (mm2) Cardiomyocyte length (mm) Nuclear area (mm2) Nuclear length (mm) Median 3.3 367.75 86.35 14.56 13.33 Mean 3.3 396 87 15 14 Standard deviation 1.1 184 26 5 3 Median 22 599.49 139.13 26.16 14.06 BDCM group Mean 22 673 142 28 14 Standard deviation 4.1 429 39 11 3

Bovine Dilated Cardiomyopathy

243

Table 2 Percentage of measurements from cases of BDCM with higher values than the maximum value obtained from the control group Number of measurements that exceed highest value in control group 0 42 70 100 45 Total number of measurements in BDCM group 270 270 480 270 45

Parameter Nuclear length (mm) Nuclear area (mm2) Cardiomyocyte area (mm2) Cardiomyocyte length (mm) Fibrosis (%)

Percentage (%) 0 16 29 37 100

A value of 0% means that both groups overlap completely; a value of 100% means no overlap.

Table 3 Estimated number of myocytes contained in the hearts of BDCM-affected and control cattle Animal no. 2* 3* 4* 6* 11 12 13 15 Myocyte length (mm) 153 150 127 152 86 73 98 78 Myocyte area (mm2) 695 377 621 393 247 340 341 513 Myocyte volume (ml3) 107 57 79 60 21 25 34 40 Fibrosis (%) 23 22 23 18 04 02 03 03 Heart volume (ml) 3868 3208 3113 2453 3302 2594 3538 2358 Number of myocytes 0.45 1012 0.69 1012 0.48 1012 0.49 1012 1.61 1012 1.07 1012 1.08 1012 0.60 1012

Mean values for numbers of myocytes in BDCM-affected and control animals (0.53 1012 and 1.09 1012, respectively) indicate a reduction in BDCM animals to 48.62% of control animals, i.e., a loss of 51.38%. *BDCM-affected. Control.

without clinical disease exhibit myocardial brosis or cardiomyocyte elongation, as found in healthy relatives of human DCM patients (Mahon et al., 2002). If this were the case, a morphometric approach based on cardiac biopsies or postmortem material would be a useful adjunct for the detection of carriers. Three uncharacterized proteins, detected in animals with BDCM, may prove to be of value as markers for the disease in genetically predisposed cattle identied through family pedigrees (Weekes et al., 1999), but as yet no direct genetic marker for the disease is available. Animals suffering from BDCM showed no increase in cardiac weight, suggesting that the increase in overall size of the BDCM heart is due to chamber dilatation rather than to increased tissue volume. This is also the case in nearly 50% of human patients (Rose and Beck, 1985; Gallo and dAmati, 2001). We conclude that the increase in chamber size observed in dilated hearts is mainly due to compensatory elongation of the cardiomyocytes and deposition of collagen in the interstitium. It is acknowledged that the pathological features measured in this study represent the terminal stages of BDCM and that it would be necessary to extend this work to animals at earlier clinical stages

to determine the rate of pathological progression. This would also enable the relevance of the different parameters to be determined and related to clinical signs. Furthermore, the morphometric techniques and statistical analysis used in this study might also be applicable to the study of DCM in other species. The extent of cardiomyocyte loss calculated in the present study indicates that the bovine myocardium has a considerable physiological reserve, and that c. 50% of cardiomyocytes can be lost before the affected animal develops clinical congestive heart failure.

Acknowledgments
We thank Drs A. Philbey and A. Russell for their critical review of this work. The archive from which the BDCM samples were obtained belongs to the Department of Veterinary Pathology and was initiated by Dr I.A.P. McCandlish. We also thank Richard Irvine and Lynn Stevenson for postmortem and histological expertise, respectively. Giles Innocent was funded by The Wellcome Trust as part of the International Partnership Research Award in Veterinary Epidemiology. This study was supported by the University of Glasgow.

244

P. Nart et al.

References
Baandrup, U., Florio, R. A., Rehahn, M., Richardson, P. J. and Olsen, E. G. (1981). Critical analysis of endomyocardial biopsies from patients suspected of having cardiomyopathy. II: Comparison of histology and clinical/haemodynamic information. British Heart Journal, 45, 487 493. Baird, J. D., Maxie, M. G., Kennedy, B. W. and Harris, D. J. (1986). Dilated (congestive) cardiomyopathy in Holstein cattle in Canada: genetic analysis of 25 cases. Proceedings of the 14th World Congress on Diseases of Cattle, Dublin, 1, 8994. Bancroft, J. D. and Stevens, A. (1996). Theory and Practice of Histological Techniques, 4th Edit., Churchill Livingstone, London. Beltrami, C. A., Della, M. V., Finato, N. and Rocco, M. (1996). Computer-assisted morphometric analysis of the heart. Annals of Quantitative Cytology and Histology, 18, 129 136. Beltrami, C. A., Finato, N., Rocco, M., Feruglio, G. A., Puricelli, C., Cigola, E., Sonnenblick, E. H., Olivetti, G. and Anversa, P. (1995). The cellular basis of dilated cardiomyopathy in humans. Journal of Molecular and Cellular Cardiology, 27, 291305. Benjamin, I. J., Schuster, E. H. and Bulkley, B. H. (1981). Cardiac hypertrophy in idiopathic dilated congestive cardiomyopathy: a clinicopathologic study. Circulation, 64, 442447. Bradley, R., Jefferies, A., Jackson, P. and Wijeratne, W. (1991). Cardiomyopathy in adult Holstein Friesian cattle in Britain. Journal of Comparative Pathology, 104, 101 112. Dick, M., Donald, V., Unverferth, M. and Baba, N. (1982). The pattern of myocardial degeneration in nonischemic congestive cardiomyopathy. Human Pathology, 13, 740 744. Dolf, G., Stricker, C., Tontis, A., Martig, J. and Gaillard, C. (1998). Evidence for autosomal recessive inheritance of a major gene for bovine dilated cardiomyopathy. Journal of Animal Science, 76, 1824 1829. Ebbeson, S. and Tang, D. (1967). A comparison of sampling procedures in a structured cell population. In: Stereology, H. Elias, Ed., Springer, New York, p. 131. Eschenhagen, T., Diederich, M., Kluge, S. H., Magnussen, O., Mene, U., Muller, F., Schmitz, W., Schlz, H., Weil, J., Sent, U., Schaad, A., Scholtysik, G., Wuthrich, A. and Gaillard, C. (1995). Bovine hereditary cardiomyopathy: an animal model of human dilated cardiomyopathy. Journal of Molecular and Cellular Cardiology, 27, 357 370. Figulla, H. R., Rahlf, G., Nieger, M., Luig, H. and Kreuzer, H. (1985). Spontaneous hemodynamic improvement or stabilization and associated biopsy ndings in patients with congestive cardiomyopathy. Circulation, 71, 1095 1104. Furuoka, H., Yagi, S., Murakami, A., Honma, A., Kobayashi, Y., Matsui, T., Miyahara, K. and Taniyama, H. (2001). Hereditary dilated cardiomyopathy in Holstein-Friesian cattle in Japan: association with

hereditary myopathy of the diaphragmatic muscles. Journal of Comparative Pathology, 125, 159 165. Gallo, P. and dAmati, G. (2001). Cardiomyopathies. In: Cardiovascular Pathology, M. D. Silver, I. G. Avrum and F. J. Schoen, Eds, Churchill Livingstone, Philadelphia, p. 286. Gerdes, A. M. (1992). Structural remodeling of cardiac myocytes in patients with ischemic cardiomyopathy. Circulation, 86, 426430. Goldstein, H. (1986). Multilevel mixed linear model analysis using iterative generalized least squares. Biometrika, 73, 43 56. Goldstein, H. (1989). Restricted unbiased iterative generalized least-squares estimation. Biometrika, 76, 622623. Ihaka, R. and Gentleman, R. (1996). R: a language for data analysis and graphics. Journal of Computational and Graphical Statistics, 5, 299 314. Junqueira, L. C., Bignolas, G. and Brentani, R. R. (1979). Picrosirius staining plus polarization microscopy: a specic method for collagen detection in tissue sections. Histochemistry Journal, 11, 447 455. Kajstura, J., Zhang, X., Liu, Y., Szoke, E., Cheng, W., Olivetti, G., Hintze, T. H. and Anversa, P. (1995). The cellular basis of pacing-induced dilated cardiomyopathy. Myocyte cell loss and myocyte cellular reactive hypertrophy. Circulation, 92, 2306 2317. Leifsson, P., Olsen, S., Agerholm, J. and Basse, A. (1994). Myocardial brosis (cardiomyopathy) in cattle. Dansk Veterinaertidsskrift, 77, 682 684. Loud, A. V. and Anversa, P. (1984). Morphometric analysis of biologic processes. Laboratory Investigation, 50, 250 261. Mahon, N. G., Madden, B. P., Caforio, A. L., Elliott, P. M., Haven, A. J., Keogh, B. E., Davies, M. J. and McKenna, W. J. (2002). Immunohistologic evidence of myocardial disease in apparently healthy relatives of patients with dilated cardiomyopathy. Journal of the American College of Cardiology, 39, 455462. Martig, J., Tschudi, P., Perritaz, C., Tontis, A. and Luginbuhl, H. (1982). Incidence of cardiac insufciency in cattle: preliminary report. Schweizer Archiv fur Tierheilkunde, 124, 69 82. McLennan, M. and Kelly, W. R. (1990). Dilated (congestive) cardiomyopathy in a Friesian heifer. Australian Veterinary Journal, 67, 7576. Mendez, J. and Keys, A. (1960). Density and composition of mammalian muscle. Metabolism, 9, 184 188. Ohtani, K., Yutani, C., Nagata, S., Koretsune, Y., Hori, M. and Kamada, T. (1995). High prevalence of atrial brosis in patients with dilated cardiomyopathy. Journal of the American College of Cardiology, 25, 1162 1169. Olsson, S. O. (1987). Cardiac insufciency in HolsteinFriesians. Svensk Veterinartidning, 39, 63 65. Paradine, C. and Rivett, B. (1960). Statistical Methods for Technologists. English Universities Press, London. Pelliccia, F., dAmati, G., Cianfrocca, C., Bernucci, P., Nigri, A., Marino, B. and Gallo, P. (1994).

Bovine Dilated Cardiomyopathy

245

Histomorphometric features predict 1-year outcome of patients with idiopathic dilated cardiomyopathy considered to be at low priority for cardiac transplantation. American Heart Journal, 128, 316 325. Porzio, S., Masseroli, M., Messori, A., Forloni, G., Olivetti, G., Jeremic, G., Riva, E., Luvara, G. and Latini, R. (1995). A simple, automatic method for morphometric analysis of the left ventricle in rats with myocardial infarction. Journal of Pharmacological and Toxicological Methods, 33, 221 229. Robinson, W. and Ferrans, V. (1975). Pathologic anatomy of the cardiomyopathies. Human Pathology, 6, 287 342. Rose, A. G. and Beck, W. (1985). Dilated (congestive) cardiomyopathy: a syndrome of severe cardiac dysfunction with remarkably few morphological features of myocardial damage. Histopathology, 9, 367 379. Rowan, R. A., Masek, M. A. and Billingham, M. E. (1988). Ultrastructural morphometric analysis of endomyocardial biopsies: idiopathic dilated cardiomyopathy, anthracycline cardiotoxicity, and normal myocardium. American Journal of Cardiovascular Pathology, 2, 137 144. Satoh, T. (1988). Studies on the dilated cardiomyopathy in cattle. Bulletin of the Nippon Veterinary and Zootechnical College, 37, 152154. Schwarz, F., Mall, G., Zebe, H., Blickle, J., Derks, H., Manthey, J. and Kubler, W. (1983). Quantitative morphologic ndings of the myocardium in idiopathic dilated cardiomyopathy. American Journal of Cardiology, 51, 501 506.

Sonoda, M., Takahashi, K., Kurosawa, T., Matsukawa, K. and Chihaya, Y. (1982). Clinical and clinicopathological studies on idiopathic congestive cardiomyopathy in cattle. In: Proceedings of the XII World Congress on Diseases of Cattle Amsterdam-Utrecht, World Association of Buiatrics, pp. 1187 1191. Spinale, F. G., Crawford, F. A. Jr, Hewett, K. W. and Carabello, B. A. (1991). Ventricular failure and cellular remodelling with chronic supraventricular tachycardia. Journal of Thoracic Cardiovascular Surgery, 102, 874 882. Strobeck, J. E. and Sonnenblick, E. H. (1986). Myocardial contractile properties and ventricular performance. In: The Heart and Cardiovascular System, H. A. Fozzard, Ed., Raven Press, New York, p. 32. Towbin, J. A. and Bowles, N. E. (2002). The failing heart. Nature, 415, 227233. Unverferth, D. V., Baker, P. B., Swift, S. E., Chaffee, R., Fetters, J. K., Uretsky, B. F., Thompson, M. E. and Leier, C. V. (1986). Extent of myocardial brosis and cellular hypertrophy in dilated cardiomyopathy. American Journal of Cardiology, 57, 816820. Weekes, J., Wheeler, C. H., Yan, J. X., Weil, J., Eschenhagen, T., Scholtysik, G. and Dunn, M. J. (1999). Bovine dilated cardiomyopathy: proteomic analysis of an animal model of human dilated cardiomyopathy. Electrophoresis, 20, 898 906.  Received; March 26th; 2003 Accepted; November 11th; 2003