Fibrous Matrix of Ventricular Myocardium in Tricuspid Atresia Compared With Normal Heart
A Quantitative Analysis
Background The collagen matrix is a small component of the myocardium, but it provides a supportive framework. An increase in collagen in the pressure-overloaded ventricle is known to cause myocardial stiffness. However, little is known about the collagen matrix in the volume-overloaded ventricle, particularly in relation to congenital heart disease.
Methods and Results We examined a total of 53 hearts with tricuspid atresia and 58 normal hearts matched for age. Using a microscopic-morphometric method, we analyzed the percentage per field area occupied by interstitial fibrous tissue in four sites in the ventricular mass for each specimen. A comparison of sampling sites showed no significant variations between normal and malformed hearts. Results from a homogeneity of regression coefficients analysis suggested that the two groups shared the same basic relation of proportion of fibrosis with age. The use of ANCOVA, however, revealed a clear separation between the extents of fibrous tissue in the two groups of hearts.
Conclusions The myocardium of hearts with tricuspid atresia is consistently more fibrotic than normal heart and is probably an inherent part of the malformation. This difference could explain, at least in part, the clinical observation that the left ventricle is frequently abnormal, even at an early age.
The collagen matrix of the myocardium, a fibrillar network, is a small constituent of the myocardial interstitium,1 but it has been recognized to provide a supportive framework for the myocardium. Although the collagenous network maintains myocytic alignment, limits slippage, and permits reversible interdigitation throughout the cardiac cycle, it also determines stiffness of the muscle.2 3 4 5 6 An abnormal accumulation of fibrous tissue in the interstitial space causes diastolic dysfunction due to an increase in myocardial stiffness as well as predisposing to abnormal electrical dispersion.7 8
Studies on the spontaneously hypertensive rat have shown, in the short term, an increase in muscle mass of the heart but, in the long term, a considerable increase in collagen.9 10 This suggests that synthesis of collagen relates to the duration of pressure overload. Other experimental studies have also shown a marked increase in myocardial collagen in the model of pressure overload but not in the system modeling volume overload.11 12 Ventricular hypertrophy of varying cause is usually, although not always, accompanied by an increase in collagen.13
Relatively little attention, however, has been paid to the functional role of increased myocardial collagen in the clinical setting, particularly in relation to congenital heart disease.14 There have been only sporadic reports on the structurally malformed heart.15 16 17 18 It may be that an abnormal deposition of fibrous tissue is a part of the natural history of some of these malformations. To investigate this hypothesis, we selected a group of hearts with tricuspid atresia. In these hearts, there is no egress from the right atrium except through a defect in the atrial septum, which allows flow to the left atrium. Systemic and pulmonary venous flows combine to enter the left ventricle. A ventricular septal defect then enables the blood to enter the right ventricle, which is rudimentary and lacks its inlet component. The systemic or pulmonary status of the dominant left and incomplete right ventricle is then determined by the ventriculoarterial connections. However, it is probably the left ventricle that has the major pumping function. Thus, this group of hearts permits investigation of whether there is an increment in fibrous tissue with age and whether there is a relation with the type of ventriculoarterial connection of the ventricle. For this study, we elected to use a microscopic-morphometric method rather than a biochemical method because microscopy has the advantage of displaying a transmural profile through the thickness of the ventricular wall and allows selection of areas without obvious scar tissue or perivascular fibrosis.
A total of 58 normal hearts and 53 hearts with tricuspid atresia were obtained from the pathological collections of the National Heart and Lung Institute (London, UK) and the Institute of Child Health (Liverpool, UK). Transmural slices of tissue were removed from four standard sites in the hearts, which had been fixed in formalin (Fig 1⇓). The sites—right ventricular apex, right ventricular outlet, left ventricular inlet, and left ventricular apex—were chosen because they were represented in hearts from both groups.
With the major descending coronary arteries used as a guide, all of the slices were taken parallel to the arteries, ie, along the long axes of the ventricles. The slices were processed routinely for histology and sectioned at 7 μm. The sections were stained in hematoxylin and eosin and with Masson's trichrome to evaluate any pathological changes. Sister sections from each slice were stained in sirius red F3BA in a supersaturated picric acid solution.
The sections stained in sirius red, a collagen-specific stain, were used for morphometric analysis. Three zones were assessed in each section. These zones (subepicardial, mesocardial, and subendocardial) were assigned arbitrarily by dividing the ventricular wall into thirds. Evaluation of fibrous tissue was made with a Leitz microscope and the Quantimat image analyzer system. The image from the microscope was transferred via a black-and-white camera to a computer monitor. The same magnification was used for each slide. The gray level of the image was adjusted on each slide through a comparison of the image viewed through the microscope with the image seen on the monitor, so that the fibrous tissue identified through red staining through the microscope appeared as black on the monitor. Care was taken to ensure that fibrous tissue was not overrepresented. The pixels of black areas were then detected with the analyzer. To avoid interobserver variations, the analyzer was operated by only one technician. The operator was blinded to the origin of the sections and the age of the patients. Because some of the heart specimens had come from patients who had undergone surgical interventions, scar tissue or replacement fibrosis was present in some sections. Also, some sections included large to medium-sized coronary arteries with thick adventitial layers. To avoid overestimation, assessment of the fields therefore excluded scar tissue and perivascular fibrous tissue. In this way, interstitial fibrosis, expressed as a percentage of the field area, was assessed in six fields in each of the three zones on each section.
Data from a pilot study of five hearts showed that it was necessary to measure the degree of fibrosis per site for each heart six times to provide a stable measure of central tendency and to support a measure of normality. Analysis of the size of samples determined that ≥50 hearts were needed in each group to be 95% confident that any differences observed were not the result of chance alone.
All patient/site/zone–specific data underwent a Shapiro-Wilk test of normality. Because sufficient data proved not to be normally distributed, we chose the median value as the most robust measure of central tendency and performed all site/zone–specific tests using this measure (or its rank) as the raw data for each patient. When site-only comparisons were made, the data used were the median value of the entire transmural region for each particular patient.
Age-matched normal control hearts were difficult to obtain and were not present in sufficient numbers. Because of this, we made an early commitment to analyze the effects of age on ventricular fibrosis in both groups of hearts. This was achieved, first, by exploring the relations between age and ventricular fibrosis on a site/zone basis using linear regression and, second, by making comparisons between the groups using an analysis of the homogeneity of regression coefficients and ANCOVA. The absolute measures of the proportion of fibrosis were then compared directly after correction for age using multivariate ANOVA.
Like the effects of age, transmural variation, as data representing the proportion of fibrosis existing for each of the three zones across the ventricle for each patient, was explored using linear regression, and comparisons were made with an analysis of the homogeneity of regression coefficients and ANCOVA after arbitrary coding of the subendocardial zone as “zone 1,” the mesocardial zone as “zone 2,” and the subepicardial zone as “zone 3.” Comparisons of the absolute proportion of fibrosis between sites were made with one-way ANOVA.
A comparison was made of variability between the sites of measurement within each group using one-way ANOVA. For the hearts with tricuspid atresia alone, comparison of the proportion of fibrosis was made between the types of ventriculoarterial connection using two-way ANOVA after we first established that ages were comparable and that therefore there was no need for age correction.
Throughout the study, probability values of P≤.05 were taken as evidence of an effect unlikely to be due to chance alone.
The study included 58 structurally normal hearts from children aged 3 days to 16 years. The causes of death were malignancy (18), cranial lesions (12), trauma (9), infection (5), gastrointestinal lesions (4), hypoglycemia (1), liver necrosis (1), uremia (1), and unknown (7). Pathological changes were not detected in any of the histological sections from these cases.
The 53 heart specimens of tricuspid atresia came from patients with an age range from neonate to 17 years. All except 2 specimens exhibited classic tricuspid atresia, ie, absence of the right atrioventricular connection with the left atrium connected to a dominant left ventricle. The rudimentary right ventricle was anterior and to the right in 48 hearts and was directly anterior in 3 hearts. One of the 2 remaining hearts had a double-inlet atrioventricular connection to a dominant left ventricle associated with an imperforate right atrioventricular valve. In the remaining heart, there were concordant atrioventricular connections, but the tricuspid valve was imperforate. All specimens had a deficient atrial septum (4 specimens had atrial septectomy), which had allowed systemic venous return to enter the left atrium. The ventriculoarterial connections were concordant in 37 hearts and discordant in 16. Aortic coarctation was present in 5 hearts, all with the aorta originating from the rudimentary right ventricle (discordant ventriculoarterial connections) and associated with a restrictive ventricular septal defect. Stenosis of the pulmonary valve was found in 3 hearts, and obstruction to pulmonary outflow due to a restrictive ventricular septal defect was found in another 3 hearts.
Signs of surgical intervention were found in 28 hearts with tricuspid atresia. Of these, 15 showed evidence of palliative procedures involving construction of a form of Blalock-Taussig shunt, atrial septectomy, or repair of coarctation. The donors of the remaining 13 hearts had undergone a modified Fontan procedure or a bidirectional Glenn procedure during life. In the latter group, the longest survival had been 4.5 months, whereas the other patients had died within 1 week of the operative procedure.
Quantification of Fibrous Tissue
Comparison of Normal Heart and Tricuspid Atresia With Age
ANCOVA for each site and zone revealed a significant difference (all P<.0001) between the extent of fibrosis for the two groups of hearts, which was consistent across the age ranges studied. The homogeneity of regression coefficients (range, .07 to .96) suggests that normal hearts and those with tricuspid atresia share the same basic relation of the proportion of fibrosis with age.
Comparison of Age and Fibrosis in Various Zones
Because regression coefficients are statistically equivalent for the two groups of hearts, regression of all data for age (from both groups of patients, corrected for differences due to diagnosis) against the proportion of fibrosis revealed a significant relation only for the subepicardial zones of all four sites (P<.05 to <.005).
A comparison of these data, stratified by age, emphasizes the reasonably consistent nature of the differences between the groups of normal heart and tricuspid atresia for zones other than the subepicardial ones (Fig 2⇓), which also appeared to be more fibrotic.
Comparison of Sites and Zones of Sampling
The magnitude of the difference in the median values between the normal and abnormal groups was highly statistically significant (all P<.0001) after correction for age using multivariate ANOVA. Again, the subepicardial zone was consistently more fibrotic in both normal and abnormal hearts in all sites sampled.
Transmural Distribution of Fibrous Tissue
Table 1⇓ shows the results of an analysis of the variation in percentage of fibrosis across the ventricular wall at each site. Visually, the regression coefficients for normal hearts and those with tricuspid atresia are very similar. This is confirmed by the nonsignificant results yielded from the analysis concerning homogeneity of regression coefficients. For each regression relation, the coefficient is significantly different from zero, suggesting an increasing proportion of fibrous tissue from the subendocardium to the subepicardium (zone 1 to zone 3). Similar to the analysis of age, the data from each site demonstrated clear differences between normal hearts and those with tricuspid atresia. Unlike with age (other than at the subepicardial zone), the combination of the two sets of data, normal and tricuspid atresia (corrected for diagnosis), demonstrated a highly significant degree of variation across the ventricular wall.
The probability matrices resulting from a one-way ANOVA with adjusted pairwise comparisons of the site-specific data for percentage of fibrosis for each group of hearts achieved significance in nearly all sites. Only the mesocardial and subendocardial zones in normal hearts at the right ventricular outlet were not significantly different. Thus, a transmural variation in the density of fibrous tissue is clearly evident in most sites in both groups.
Comparison of Hearts With Concordant and Discordant Ventriculoarterial Connections
The age at death was not significantly different between the two groups (concordant, newborn to 17 years; discordant, 3 days to 14.5 years). A tendency was found toward a higher percentage of fibrous tissue in all sites in hearts with concordant ventriculoarterial connections (Table 2⇓). There was no difference in percentage of fibrosis between the left ventricle supporting the systemic circulation (concordant ventriculoarterial connections) and the left ventricle supporting the pulmonary circulation (discordant ventriculoarterial connections). Similarly, there was no difference between the right ventricle supporting the aorta or that supporting the pulmonary trunk. Table 3⇓ shows that the type of ventriculoarterial connections and the site of sampling either as independent variables or taken together (interaction) did not affect the proportion of fibrous tissue.
Studies on the fibrous network supporting the ventricular myocardium have focused mainly on conditions of pressure overload, with conditions involving volume overload and congenital malformations receiving little attention. With experimentally induced volume overload, Ruzika et al19 found a decrease in left ventricular collagen compared with control hearts. In contrast, the right ventricle was found to have an increased content of collagen only during the initial weeks after the construction of an aortocaval shunt. In another study of volume overload, in dogs, Iimoto et al20 also found a decrease in the major types (I and III) of collagen. Both studies used analysis of hydroxyproline as a measure of collagen. Analysis using cyanogen bromide peptide substantiated the hypothesis of Iimoto et al20 that there was an increase in cross-linking, thus resulting in a decrease in the collagen that could be detected through hydroxyline proline analysis. Our study, in which we used a morphometric method, did not distinguish between the types of collagen but measured all collagen detectable by using sirius red.21 Our results, as judged by the homogeneity of regression coefficients, suggest that normal and abnormal hearts have the same basic relation of percentage of fibrous tissue per field, but the abnormal hearts consistently have a higher proportion than normal (Fig 2⇑). Because an increase in ventricular mass in tricuspid atresia is well recognized,22 the absolute increase in fibrous content is likely to be higher. An increase in myocytic size would have a diluting effect on the percentage of fibrous tissue per field. Any differences we demonstrated, therefore, are likely to be less than the true values.
Neither group of hearts showed a difference in fibrous tissue with increasing age other than in the subepicardial zones. Again, because we have not taken into account the growth of the myocytes, any increase may well have been eclipsed. Our observation concurs with two studies on hearts from juveniles and adults,13 23 although another study24 showed an increase in the total amount of collagen with age. When we analyzed our data for variation between sites, the analysis did not achieve statistical significance.
Our analysis for comparing the type of ventriculoarterial connections in tricuspid atresia did not show any significant difference in the proportion of fibrous tissue. This probably reflects the dominant role of the left ventricle, with the right ventricle being functionally an extension of the left ventricle. Alternatively, the left ventricle has the same role regardless of the type of ventriculoarterial connections because it is the ventricle that distributes the entire cardiac output. Our number of hearts with obstruction to egress to either great artery was too small to allow an investigation of the effect of stenosis of outflow tracts on the proportion of fibrous tissue.
Clinical studies have implicated inappropriate left ventricular hypertrophy, and subsequent myocardial hypoxia leading to fibrosis in the subendocardium, as the major determinant of functional deterioration.25 Our study showed transmural variation at all sites, but we observed more fibrous tissue in the subepicardial than in the subendocardial region. This finding is at variance with other studies.26 27 The gradient of variation was a common feature in both our normal and abnormal series. The diversity in results from the various studies, therefore, is likely to be related to differences in methodology and orientation of the sections sampled. Our results, nevertheless, suggest that the increase in fibrous tissue seen in the malformed heart is an inherent component of the lesion affecting both right and left ventricles. The difference when compared with normal hearts is observed even in the youngest patients of our series. This difference could account, at least in part, for the impairment of left ventricular function observed from an early age in patients with tricuspid atresia,22 supporting the notion of Baker et al28 that the dominant left ventricle may be intrinsically abnormal.
We, of course, were not able to exactly match the two groups of hearts by age, but this problem has largely been circumvented by correcting for the age difference in the statistical analysis. Our analysis has shown that age is not a major factor as it affects only the subepicardial zones. The difference between normal and tricuspid atresia is apparent even in the raw data uncorrected for age. Furthermore, by analyzing percentage per field area, without correcting for growth or hypertrophy, we consistently underestimated the amount of fibrous tissue present. Finally, we analyzed only postnatal cases. To understand whether the increase in fibrous tissue found in tricuspid atresia is truly part of the heart defect, the study would have to include fetal cases. Again, the availability of materials would be limited.
This study was supported by a grant from the British Heart Foundation (PG93060). The authors thank Dr Robert Anderson for reviewing the manuscript in its draft form.
- Received December 20, 1995.
- Revision received March 13, 1996.
- Accepted April 11, 1996.
- Copyright © 1996 by American Heart Association
Frank JS, Larger GA. The myocardial interstitium: its structure and its role in ionic exchange. J Cell Biol. 1974:60;586-601.
Fry DL, Griggs DM, Greenfield JC. Myocardial mechanics: tension-velocity length relationships of heart muscle. Circ Res. 1964;14:73-85.
Weber KT, Brilla CG, Janicki JS. Myocardial fibrosis: functional significance and regulatory factors. Cardiovasc Res. 1993;27:341-348.
Weber KT, Janicki JS, Shroff SG. Collagen compartment remodelling in the pressure overloaded left ventricle. J Appl Cardiol. 1988;1:37-46.
Daliento L, Scognomiglio R, Thiene G, Hegerty A, Ho SY, Caneve F, Anderson RH. Morphologic and functional analysis of myocardial status in pulmonary atresia with intact ventricular septum: an angiographic, histologic and morphometric study. Cardiol Young. 1992;2:361-366.
Jones M, Ferrans VJ, Morrow AG, Roberts WC. Ultrastructure of crista supraventricularis muscle in patients with congenital heart diseases associated with right ventricular outflow tract obstruction. Circulation. 1975;51:39-67.
Hess OM, Schneider J, Koch R, Bamert C, Grim J, Krayenbuehl HP. Diastolic function and myocardial structure in patients with myocardial hypertrophy: special reference to normalized viscoelastic data. Circulation. 1981;63:360-371.
Krayenbuehl HP, Hess OM, Monrad SE, Schneider J, Mall G, Turina M. Left ventricular myocardial structure in aortic valve disease before, intermediate, and late after aortic valve replacement. Circulation. 1989;79:744-755.
Ruzika M, Keely FW, Leenen FHH. The renin-angiotensin system and volume overload-induced changes in cardiac collagen and elastin. Circulation. 1994;90:1989-1996.
Iimoto DS, Covell JW, Harper E. Increase in cross-linking of type I and type III collagens associated with volume-overload hypertrophy. Circ Res. 1988;63:399-408.
Fuster V, Danielson MA, Robb RA, Broadbent JC, Brown AL, Elveback LR. Quantitation of left ventricular myocardial fiber hypertrophy and interstitial tissue in human hearts with chronically increased volume and pressure overload. Circulation. 1977;55:504-508.
Baker EJ, Jones ODH, Joseph MC, Maisey MN, Tynan MJ. Radionuclide measurement of left ventricular ejection fraction in tricuspid atresia. Br Heart J. 1984;52:572-574.