This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Koide, M. Articles by Carabello, B. A. Search for Related Content PubMed PubMed Citation Articles by Koide, M. Articles by Carabello, B. A.

(Circulation. 1997;95:1601-1610.)
© 1997 American Heart Association, Inc.

Articles

Premorbid Determinants of Left Ventricular Dysfunction in a Novel Model of Gradually Induced Pressure Overload in the Adult Canine

Masaaki Koide, MD; Masayoshi Nagatsu, MD; Michael R. Zile, MD; Masayoshi Hamawaki, MD; M. Michael Swindle, DVM; George Keech, VMD; Gilberto DeFreyte, BS; Hirofumi Tagawa, MD; George Cooper, IV, MD; Blase A. Carabello, MD

From the Cardiology Division and Gazes Cardiac Research Institute (M.K., M.N., M.R.Z., G.D., H.T., G.C., B.A.C.) and the Department of Comparative Medicine (M.M.S., G.K.), Medical University of South Carolina, Charleston, and the Ralph H. Johnson Department of Veterans Affairs (M.R.Z., G.C., B.A.C.), Charleston, SC.

Correspondence to Blase A. Carabello, MD, Cardiology Division, Medical University of South Carolina, 171 Ashley Ave, Charleston, SC 29425-2221.

	Abstract

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
Background When a pressure overload is placed on the left ventricle, some patients develop relatively modest hypertrophy whereas others develop extensive hypertrophy. Likewise, the occurrence of contractile dysfunction also is variable. The cause of this heterogeneity is not well understood.

Methods and Results We recently developed a model of gradual proximal aortic constriction in the adult canine that mimicked the heterogeneity of the hypertrophic response seen in humans. We hypothesized that differences in outcome were related to differences present before banding. Fifteen animals were studied initially. Ten developed left ventricular dysfunction (dys group). Five dogs maintained normal function (nl group). At baseline, the nl group had a lower mean systolic wall stress (96±9 kdyne/cm²; dys group, 156±7 kdyne/cm²; P<.0002) and greater relative left ventricular mass (left ventricular weight [g]/body wt [kg], 5.1±0.36; dys group, 3.9±0.26; P<.02). On the basis of differences in mean systolic wall stress at baseline, we predicted outcome in the next 28 dogs by using a cutoff of 115 kdyne/cm². Eighteen of 20 dogs with baseline mean systolic stress >115 kdyne/cm² developed dysfunction whereas 6 of 8 dogs with resting stress 115 kdyne/cm² maintained normal function.

Conclusions We conclude that this canine model mimicked the heterogeneous hypertrophic response seen in humans. In the group that eventually developed dysfunction there was less cardiac mass despite 60% higher wall stress at baseline, suggesting a different set point for regulating myocardial growth in the two groups.

Key Words: hypertrophy • stenosis • ventricles • aorta • heart failure • valves

	Introduction

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
When a pressure overload is placed on the myocardium, concentric hypertrophy develops. Hypertrophy is an important compensatory mechanism because it normalizes wall stress.¹ That load is a prime regulator of muscle mass was supported by Cooper et al.² In these experiments, right ventricular pressure overload was created in cats by pulmonary artery banding. However, at the same time a papillary muscle was unloaded in isolation by transection of the chordae tendineae. Despite the presence of overall ventricular pressure overload, the papillary muscle atrophied, demonstrating the importance of local loading in regulating muscle mass.

If the amount of left ventricular hypertrophy that developed responded to load perfectly so that wall stress was normalized, then there would be just enough increase in wall thickness to offset the increase in pressure. However, in many instances, the hypertrophy that develops is inadequate to maintain normal wall stress; wall stress increases and ejection performance falls.^{3 4} In some such cases, contractile dysfunction develops (pathological hypertrophy) in addition to afterload being increased.⁴ In other cases there is "excessive" hypertrophy producing subnormal wall stress and supernormal ejection performance.^{5 6}

The question addressed in this study is, why does a pressure overload cause the disparate myocardial responses noted above? Are these differences in response based on differences in the load—its magnitude, time course, duration, etc—or are these differences predicated on inherent differences in the myocardial response to the overload? Is there a different set point or threshold for stress to activate the hypertrophic process? In humans, years of observation of a large number of patients would be required to correlate differences in the disease process with differences in outcome with regard to myocardial hypertrophy in an effort to answer the questions posed above. However, we recently developed a model of gradual ascending aortic constriction in the adult dog. In this model, the band is applied identically to each subject, making it unlikely that differences in outcome are due to differences in the "disease" process. In our early experience with this model, we noted that similar to humans, some mongrel outbred purpose-bred dogs developed extensive left ventricular hypertrophy with normal or even supernormal left ventricular performance while other animals of similar weight, sex, and breeding developed less extensive hypertrophy and subsequently acquired left ventricular dysfunction. Since the banding technique and pressure gradient were nearly identical from animal to animal, it suggested that differences in outcome were predicated on individual differences in the response of the myocardium to the pressure overload. If that were so, we hypothesized that there might already be individual differences in the myocardial response to the existing load in the normal state before banding. If so, those differences, when found, should make the eventual outcome of banding predictable. We tested this hypothesis by retrospectively examining the baseline data for animals for which the outcome was known and then validated the importance of the differences found by using them to predict outcome in a second prospectively analyzed series of dogs.

	Methods

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
Study Design
Retrospective Studies
Fifteen parasite-free adult mongrel dogs (13 males and 2 females weighing 20 to 29 kg; age, 1 to 5 years) were studied longitudinally from baseline (when they were normal) through 8 weeks of gradually imposed ascending aortic constriction with the use of a novel banding technique. Hemodynamics and left ventricular mechanics were studied at cardiac catheterization 1 to 2 weeks before the aortic banding operation. During banding, an initial gradient of 30 mm Hg was created. At 2, 4, and 6 weeks after banding, hemodynamics and left ventricular mechanics were examined at cardiac catheterization; then the pressure overload was increased by tightening the band. Thus, the gradient was increased four times in all, and the dogs were studied at baseline and at 2, 4, 6, and 8 weeks. Three days after the final study, the dogs were humanely killed, at which time left ventricular cardiocytes were obtained for examination of cellular contractility. Cardiocyte studies were performed in 6 of the 10 dogs that developed dysfunction and 4 of the 5 dogs with normal function. As in previous studies, cardiocyte mechanics was assessed by investigators at our institution (H.T., G.C.) blinded to the in vivo studies of contractility.^{7 8} The dogs were then grouped according to the presence or absence of in vivo ventricular dysfunction, which was corroborated by the cardiocyte studies in each of the 10 cases for which these data were also available.

Prospective Studies
Differences between the group of dogs with ventricular dysfunction and the group with normal function were then used to prospectively predict functional outcome in 28 additional dogs. Thus, in all, 43 dogs were studied.

At all times the standards of care for the animals met or exceeded those of the American Physiological Society and the American Association for Accreditation of Laboratory Animal Care.

Assessment of Contractile Function
Hemodynamic studies were performed in both the ß-blocked and unblocked state at all five observations. ß-Blockade was used so that innate contractile function could be studied in the absence of adrenergically mediated changes in contractility. Unblocked hemodynamics was studied to better estimate ambient hemodynamic conditions such as gradient. Thus, when we report ventricular performance data (mean normalized ejection rate, ejection fraction, etc) and wall stress used in evaluating contractile function, these data were obtained on ß-blockade. All other data are from measurements made off ß-blockade. Contractile function was assessed by plotting mean normalized systolic ejection rate (MNSER)⁹ against mean systolic midwall stress (MSS), using data acquired during ß-blockade. This relationship is a modification of the technique that plots the mean velocity of circumferential fiber shortening (Vcf) against wall stress.¹⁰ Mean normalized systolic ejection rate is ejection fraction divided by ejection time, whereas mean velocity of circumferential fiber shortening is shortening fraction divided by ejection time. Both use a relatively preload-independent parameter (MNSER or Vcf)¹¹ and normalize it for afterload, yielding a measure of contractile function. The MNSER-MSS relation and its 95% confidence limits were plotted for 40 normal dogs previously studied on ß-blockade in our laboratory. Animals in the current study falling down and to the left of the lower confidence limit were defined as having systolic contractile dysfunction. Because MNSER reflects endocardial shortening, which might overestimate total myocardial performance,¹² modified midwall fiber shortening rate was also plotted against mean systolic midwall stress.

Model for Inducing Gradual Aortic Constriction
Since sudden pressure overload may cause acute myocardial damage¹³ and since most human pressure overloads develop gradually, it is preferable that pressure overload models induce the overload gradually. Many previous models produced gradual constriction by banding the pulmonary artery or aorta in juvenile animals. Subsequent growth produced gradually increased pressure overload as cardiac output increased through the fixed obstruction. However, juvenile hypertrophy may differ from adult hypertrophy.¹⁴ In this study, adult dogs were banded by the investigators who perfected the model (M.K., M.N.). Anesthesia was induced with an intravenous injection of a mixture of fentanyl and droperidol (0.5 mL/kg). The dogs then were intubated and mechanically ventilated. Anesthesia was maintained with inhalation of a mixture of 1% to 1.5% isoflurane, nitrous oxide, and oxygen. Two 5.2F pigtail catheters were inserted into the left femoral artery and advanced to monitor left ventricular pressure and descending aortic pressure simultaneously. A right thoracotomy was made through the third intercostal space. The pericardium was opened and the ascending aorta was exposed. To provide exposure of the aortic root, the right atrial appendage was retracted caudally. We used extreme care to avoid vasa vasoral damage; periaortic fat was dissected posteriorly 1 cm below the subclavian artery to make a small tunnel for passage of the band. A PTFE tube (6-mm diameter, Impra Inc) was placed around the ascending aorta through the tunnel and then two 9-mm-wide polyester umbilical tapes (Nittyo Kogyo Inc) were threaded through the PTFE tube. The PTFE tube protected the aortic wall from the edges of the band, which could have been a focal point for aortic rupture. The band was then inserted through a silicone stent and tied around a balloon dilatation catheter so that inflation of the catheter tightened the band, in turn constricting the aorta (Fig 1). The stent was separated from the aorta by a convex silicone "protector," which allowed the force of the constricting band to be diffused over a large area of the aorta instead of concentrated at the edges of the stent. Mattress sutures of 3-0 prolene were placed on the band ends to fix them around the balloon. The balloon was inflated with saline to test the system and to produce a peak pressure gradient across the band of 30 mm Hg. The inflation port of the balloon catheter was sealed with a rubber cap and placed in a subcutaneous pocket on the neck or back. The chest was closed, and the dogs recovered from anesthesia and were followed on a standard diet under close veterinary supervision.

View larger version (123K): [in this window] [in a new window]   Figure 1. Banding apparatus is demonstrated.

Catheterization Protocol
All five studies were performed during light anesthesia, using a mixture of fentanyl-droperidol given intravenously and inhalation of nitrous oxide–oxygen at a ratio of 3:1. This combination has been demonstrated to have little effect on contractile function.¹⁵ At the baseline study, a cutdown was performed over the carotid artery from which catheters were introduced into the left ventricle and ascending aorta. At 2, 4, and 6 weeks and at the final study, catheters were introduced into the left ventricle and aorta through a femoral artery cutdown. The studies were performed identically at each time period. Left ventricular pressure and aortic pressure were recorded simultaneously while a left ventriculogram was performed in the 30° right anterior oblique position at 60 frames per second by injection of nonionic radiographic contrast. Then, ß-blockade was induced by infusion of a loading dose of esmolol (0.5 mg/kg per minute for 3 minutes) followed by a constant infusion at a dose of 0.3 mg/kg per minute. A second ventriculogram was then performed in the same fashion as the first. The ventriculograms and the pressure recordings were used to calculate ejection fraction, MNSER, and left ventricular mass, volume, and wall stress frame-by-frame.

Augmentation of Pressure Overload
Pressure overload was augmented at 2 weeks, 4 weeks, and 6 weeks of the study as follows. After discontinuation of ß-blockade followed by a 20-minute recovery period, the balloon of the aortic constriction system was inflated by injection of saline through the injection port. The balloon was inflated slowly to increase the gradient of 30 mm Hg present after banding to a target pressure gradient of 60 mm Hg at 2 weeks, 90 mm Hg at 4 weeks, and 120 mm Hg at 6 weeks. To confirm that the 2-week interval between pressure gradient increases was adequate for full development of hypertrophy, 6 study dogs underwent daily echocardiograms. Fig 2 demonstrates that left ventricular muscle mass as measured by cross-sectional area increased over the first 9 days and then plateaued. Thus, the hypertrophy of these left ventricles appeared to be completely developed before the next pressure increase occurred.

View larger version (17K): [in this window] [in a new window]   Figure 2. Time course of the development of hypertrophy (increase in cross-sectional area, CSA) after balloon inflation is shown for 6 dogs. The curve was fitted using the near-optimum piecewise linear regression equation y=6.07+1.53x-1-3.25x-9+1.82x-10, where y=increase in CSA, x=days after overload increment, x=max{0,x}. Muscle cross-sectional area began increasing by the first day after overload increment and continued to rise until the ninth day, when it plateaued.

Isolated Cardiocyte Function
One week after the final in vivo evaluation of left ventricular function, cardiocytes were isolated and cardiocyte contractile function was measured as previously described.⁸ The dogs were deeply anesthetized. The pericardium was excised, and the heart was rapidly removed and placed in a cold calcium-free buffer. A wedge of the left ventricle supplied by the circumflex artery was isolated, and the artery was cannulated and perfused with collagenase. Then the tissue was minced into 2-mm cubes and was gently agitated for 5 minutes at 37°C while being gassed with 100% O₂. The cardiocytes were harvested by drawing off the supernatant in which they were suspended for filtration through 210-µm nylon mesh. Laser diffraction was then used to analyze the extent and velocity of sarcomere shortening. The cardiocytes were stimulated to contract between platinum wire electrodes. Changes in sarcomere length were measured from the movement of the first-order diffraction pattern cast by a substage laser light passing through the sarcomeres of a given cardiocyte onto diametrically opposed optical sensors situated above the microscope stage. Only cardiocytes that were single, rod-shaped, unattached to adjacent cells, and that contracted with each stimulus but were quiescent between stimuli were studied. Cardiocyte data from 8 previously unreported normal dogs served as control data.

Calculations
Left ventricular volumes were calculated by the area-length method. Left ventricular mass was calculated by the method of Rackley et al,¹⁶ which has previously been demonstrated to produce accurate volumes and masses in our laboratory.¹⁷ Our ventriculographic mass calculations were corroborated by calculations made from echocardiographic images as we have described before.¹⁸ Left ventricular wall stress () was calculated using Mirsky's formula¹⁹ and methods we have used previously.^{5 17} MSS was derived by averaging the stress calculated frame-by-frame during systole. MNSER was calculated as ejection fraction divided by ejection time. Left ventricular contractile function on ß-blockade was inferred by plotting the MNSER-MSS relationship of a given animal at a given observation period against the confidence limits of the relationship derived from 40 normal dogs studied during ß-blockade. Modified midwall fiber shortening velocity (midwall mVcf) was calculated as previously described¹² (see "Appendix").

Statistics
All the data are expressed as SEM. The normal MNSER-MSS and modified midwall Vcf-MSS relationships were derived using least-squares linear regression analysis. For the analysis of daily increase in the cross-sectional area of the left ventricle, piecewise linear regression²⁰ was used to search for changes in the rate of development of hypertrophy. Comparison of differences in sarcomere shortening velocity among the three groups was analyzed with ANOVA, and a Newman-Keuls test was used to determine individual differences. When grouped baseline data were compared, an unpaired t test was used. Comparisons of overall differences between the two groups over time were analyzed with ANCOVA.

	Results

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
Classification by Contractile Function (Retrospective Studies)
Two of the 17 dogs suffered aortic rupture. The remaining 15 dogs were entered into the study. Ten dogs eventually developed systolic dysfunction (dys group) at the final study as determined by the MNSER-MSS relationship obtained during ß-blockade (Fig 3, top) and confirmed by the Vcf_MW-MSS relationship (Fig 3, bottom). Five dogs maintained normal systolic function (nl group). The figures demonstrate a separation between the two groups even at baseline. Of the two female dogs studied, one resided in each group. There were no differences in weight or age between the groups. Fig 4 demonstrates the contractile function of isolated cardiocytes from the ventricles of both groups of dogs. Fig 4 (top) shows that mean sarcomere shortening velocity was less for cardiocytes taken from dogs with in vivo left ventricular dysfunction than it was in cardiocytes from controls or for cardiocytes taken from dogs with hypertrophy and normal function. Fig 4 (bottom) shows the correlation between ventricular systolic ejection rate and cardiocyte shortening velocity. Although MNSER occurs in the presence of afterload and sarcomere shortening velocity is an unloaded measurement, there is still a reasonable correlation between the in vivo and in vitro measurement of contractile function. Baseline comparisons between the two groups of dogs that eventually had dysfunction versus normal function at the studies' end are demonstrated in Table 1. Significant differences existed in ejection fraction, systolic wall stress, diastolic wall stress, and normalized left ventricular mass at baseline. However, stepwise multivariate analysis found wall stress as the only significant predictor of outcome (P<.01).

View larger version (23K): [in this window] [in a new window]   Figure 3. Top, The 15 dogs studied retrospectively are plotted against the relationship of mean normalized systolic ejection rate and mean systolic wall stress obtained from 40 normal dogs together with the 95% confidence bands of the relationship. Although both groups began with normal contractile function using this paradigm, the group that eventually developed dysfunction (dys) started out with higher stress and lower shortening rate than the group that maintained normal function (nl). Bottom, The 15 dogs are plotted against the relationship of the rate of modified midwall shortening (midwall mVcf) and mean midwall systolic stress obtained from the same normal dogs. The results using the second method of evaluation are concordant with those shown in the top plot.

View larger version (17K): [in this window] [in a new window]   Figure 4. Top, Sarcomere shortening velocity of cardiocytes taken from 6 dogs with hypertrophy and ventricular dysfunction (, dys), 4 dogs with hypertrophy and normal function (, nl), and that of 8 control dogs (, control) is shown at the end of the study. Sarcomere shortening velocity was depressed in cardiocytes taken from the dogs with ventricular dysfunction, confirming our in vivo results. Bottom, In vitro cardiocyte sarcomere shortening velocity is plotted against in vivo mean normalized systolic ejection rate for the subjects shown in the top plot. Despite differences in loading conditions, there is reasonable correlation (r=.82) between the in vitro and in vivo measures of shortening rate.

View this table: [in this window] [in a new window]   Table 1. Hemodynamic and Volumetric Data at Baseline

Left Ventricular Wall Stress, Performance, and Hypertrophy
The transconstriction pressure gradient recorded off ß-blockade at each observation period is demonstrated for both groups of dogs in Fig 5. Mean normalized systolic ejection rate on ß-blockade is plotted for the entire course of the study in Fig 6. It was maintained at baseline values in the normal group throughout the study but declined sharply at 6 weeks in the dysfunction group. Likewise, ejection fraction declined from 0.58±0.02 to 0.51±0.03 in the dysfunction group (P<.05) but was maintained in the normal group (0.72±0.03, baseline; 0.77±0.03, final study). Mean left ventricular systolic wall stress calculated from measurements made off ß-blockade and left ventricular mass are demonstrated for both groups of dogs throughout the course of the study in Fig 7. In the group that maintained normal function, left ventricular wall stress was less and mass was greater than in the group that developed dysfunction. Gross pathology sections from representatives of each group demonstrating the different patterns of hypertrophy are shown in Fig 8.

View larger version (16K): [in this window] [in a new window]   Figure 5. Transband pressure gradient throughout the course of the study was not different between the two groups, indicating relatively similar overload in both groups. See Fig 4 for abbreviations and symbols.

View larger version (18K): [in this window] [in a new window]   Figure 6. Mean normalized systolic ejection rate (MNSER) obtained during ß-blockade is shown for both groups throughout the course of the study. The course of this parameter over time was different for the two groups. In general, MNSER was less in the group that eventually developed dysfunction. Until the 6-week time period, this reduction in MNSER can be explained by the higher wall stress in the dysfunction group. However, after 6 weeks there was an abrupt fall in MNSER out of proportion to increased wall stress, indicating that contractile dysfunction had developed, as shown in Fig 3. See Fig 4 for abbreviations and symbols.

View larger version (19K): [in this window] [in a new window]   Figure 7. Top, Mean systolic stress obtained off ß-blockade is plotted for both groups throughout the study. The group that maintained normal function (, nl) had lower wall stress than the group that developed dysfunction (, dys). Bottom, Left ventricular mass (LVM) to body weight (BW) ratio, an index of hypertrophy, is plotted for both groups throughout the study. Although mass increased in both groups, the course of the development of hypertrophy was different for the two groups.

View larger version (122K): [in this window] [in a new window]   Figure 8. Left, Representative left ventricular anatomical cross section and left ventricular and aortic pressure tracings from a dog with normal contractile function. Right, Cross section and pressure tracings from a dog with dysfunction. Despite similar pressure gradients, there are obvious differences in the extent of hypertrophy.

Prospective Studies
Of the differences between the groups of dogs shown in Table 1, the difference in mean systolic wall stress was striking and statistically the most significant. After analyzing initial mean stress off ß-blockade, we picked a cutoff value for mean stress of 115 dyne/cm²x10³. Eighteen of 20 dogs with a baseline MSS of 115 kdyne/cm² developed dysfunction, whereas 6 of 8 dogs with a baseline MSS 115 kdyne/cm² had normal function. The baseline MSS of all 43 dogs in relationship to left ventricular dysfunction at 8 weeks is shown in Fig 9. As can be seen, 11 of 13 dogs with MSS 115 kdyne/cm² had normal function, whereas 28 of 30 dogs with MSS >115 kdyne/cm² developed dysfunction.

View larger version (14K): [in this window] [in a new window]   Figure 9. Baseline mean systolic wall stress (off ß-blockade) is plotted against eventual left ventricular function at the end of the study for all 43 dogs. The groups are separated by a stress of 115 kdyne/cm2 (horizontal line) except for 4 dogs (2 in each group) that responded in an opposite fashion to that predicted by baseline stress.

The baseline differences in stress and left ventricular mass between the groups might have biased the outcome in a self-fulfilling prophecy. That is, the lower mass and higher stress in the high stress group may have created an automatic "disadvantage," causing these hearts to develop failure because they began the study with lower mass to begin with. To address this problem, we banded 3 dogs with high stress (in which ventricular dysfunction was predicted to occur) to a slightly higher initial gradient (50 mm Hg) than the usual protocol. Within 1 week, left ventricular mass–to–body weight ratio had increased to 4.7±0.3, not different from that of the normal function group at baseline. Left ventricular function was normal at that time. The gradient was then decreased so that wall stress fell to that of the normal function group and remained so at 2 weeks (Fig 10). Thus, at this point in time, this manipulated group of 3 dogs initially predicted to develop dysfunction now had wall stress and left ventricular mass similar to the group that eventually maintained normal function. From there, the protocol was followed in normal fashion. By 8 weeks, wall stress in the manipulated group increased to that of the initial dysfunction group, and all 3 stress-manipulated dogs developed left ventricular dysfunction.

View larger version (20K): [in this window] [in a new window]   Figure 10. Systolic stress is shown for dogs with normal ventricular function and those with dysfunction taken from Fig 7. Superimposed are results of 3 animals with initial high stress converted to low stress via balloon manipulation by 2 weeks into the study. Despite the resultant low stress at 2 weeks in this group, high stress and ventricular dysfunction developed when the rest of the protocol was carried out.

Baseline Reproducibility
Because our study relied heavily on differences detected at baseline, we repeated the baseline study 1 week later in 14 dogs and compared the two baseline results (Table 2). Selected parameters are demonstrated in Table 2. Naturally, some variation existed, but the differences were neither clinically nor statistically significant.

View this table: [in this window] [in a new window]   Table 2. Dual Baseline Observations

	Discussion

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
In humans, the hypertrophic response to a pressure overload is heterogeneous. Extensive hypertrophy develops in some patients, whereas less hypertrophy develops in others.^{3 4 5 6} These differences could be due to differences in the disease process or to more general differences in the regulation of myocardial mass in response to a given load. We reproduced this heterogeneity of hypertrophic response in a canine model. A striking finding in this study was the clear baseline differences in left ventricular wall stress and left ventricular muscle mass that existed between animals that developed hypertrophy and eventual muscle dysfunction and animals that developed more extensive hypertrophy but maintained normal muscle function. Pressure-overloaded animals that developed muscle dysfunction had higher systolic wall stress and lower muscle mass (within the previously reported normal range)¹² before banding than animals that maintained normal muscle function. We recognize that by placing a pressure overload of equal magnitude on ventricles with less mass, wall stress increased more in that group, a factor that probably helped cause the eventual muscle dysfunction that occurred as the ventricles were unable to accommodate the increased load. Indeed, this is an important aspect of the model that may offer an insight into how similar degrees of aortic stenosis severity produce such disparate results in terms of mass and function in humans. Of particular interest is that the group that eventually developed dysfunction never responded to increasing pressure load and higher stress by developing enough mass to reduce wall stress to the level of the group that maintained normal function, even though there apparently was time for additional hypertrophy to occur (Fig 2). This was true even in the group of high-stress dogs manipulated to temporarily have low stress. Since the dogs that eventually developed dysfunction had less left ventricular mass despite 50% to 60% higher stress at baseline, we speculate that the processes responsible for regulating myocardial mass had a different set point or threshold in response to wall stress or that there was a different capacity for producing hypertrophy in the two groups. These ventricular differences occurred in outbred purpose-bred dogs that were similar in age, weight, and sex to one another. Thus, it is unlikely that differences in simple developmental characteristics or differences in physical conditioning or diet could explain the baseline ventricular differences that we found. Rather, the differences in outcome with regard to left ventricular muscle mass and function may have been due to inherited differences that regulate the way the myocardium responds to load, although obviously we have presented no direct genetic evidence to support this speculation. Different responses to load could have been modulated in part by a neurohumoral system such as the adrenergic nervous system or renin-angiotensin system. Alternatively, the prebanding differences in stress and mass could have by themselves led to the differences that occurred after banding; thus the relatively higher baseline stress in the dysfunction group may have approached a hypothetical, maximally tolerable stress, prohibiting the myocardium from adapting to still higher stress, eventually producing the dysfunction. However, the results of the group of manipulated high stress dogs militate against this explanation. Dogs with initially high stress and low mass were manipulated at one point in the study to have higher mass and lower stress so that they would be similar to the normal function group, in effect giving them a new baseline. Nonetheless, they developed dysfunction, as predicted by their initially high stress.

We believe that our findings are of interest because they are so similar to those that occur in human adults who acquire aortic stenosis. Many patients demonstrate afterload mismatch, a situation in which hypertrophy is inadequate to normalize systolic wall stress; such patients often develop left ventricular muscle dysfunction.^{3 4} On the other hand, there are patients with exuberant hypertrophy, reduced wall stress, and excellent ejection performance.⁶ These patients are typified by our group of dogs that maintained normal contractile function throughout the study. Indeed, these dogs had an average ejection fraction of 77% at the end of the study.

To our knowledge this is the first time that these baseline observations and their correlation with the response to pressure overload have been made in either experimental or human aortic stenosis. In human studies of aortic stenosis, the results of the pressure overload have been examined only after the load has been well established. The reasons for this are obvious. There would be no cause to study patients prospectively at the time when they were normal, since such patients would not have come under clinical scrutiny. Likewise, in most animal studies of pressure overload, the effect of the overload was studied only after the pressure overload caused muscle dysfunction. In one longitudinal study, Sasayama and colleagues²¹ did examine the effects of modest pressure overload that produced 20% left ventricular hypertrophy and normal left ventricular function. Since dysfunction never developed, it was not possible to examine differences between animals with normal function and animals with dysfunction. Aoyagi and colleagues²² recently followed banded sheep longitudinally from when contractile function was normal through the development of intrinsic contractile dysfunction. The results of their study and ours were similar in that both found a period of compensated hypertrophy before dysfunction developed. However, our study differs from theirs in that we found two distinct groups of animals with different hypertrophic responses and different functional outcomes.

We should point out that at the time when the animals were normal, it obviously was impossible to weigh the hearts and at no time during an in vivo study was it possible to directly measure wall stress. Thus, baseline mass and stress, two major properties upon which our conclusions rest, were calculated rather than directly measured. However, we have established accurate mass calculations made from echocardiograms or ventriculograms validated by mass weighed after the animals were euthanatized in our laboratory.^{17 18} Accurate mass calculation requires accurate measurement of dimensions and thickness. Since these are the same dimensions from which wall stress is calculated (when pressure is added), accurate mass calculation helps validate our ability to calculate stress.

Extent of Hypertrophy Versus Left Ventricular Dysfunction
In previous experimental studies of pressure overload induced by outflow obstruction, there has been a rough positive correlation between the extent of hypertrophy and the development of left ventricular dysfunction, with dysfunction occurring when the hypertrophy was severe. However, a partial summary (Table 3) of studies of function in hypertrophy demonstrates many exceptions to this concept.^{21 22 23 24 25 26 27 28 29} For instance, contractile function was normal in one group of animals with 123% hypertrophy²⁶ yet abnormal in another group with less hypertrophy (45%)²⁴ but also abnormal in the group with more hypertrophy (150%).²⁹ These data suggest that the development of left ventricular dysfunction is not based simply on developing extensive hypertrophy. Indeed, in the present study, the animals with the most hypertrophy had normal function whereas those with less hypertrophy had muscle dysfunction. As noted above, these traits also may be seen in humans.^{3 4 5 6} These studies suggest that hypertrophy can be compensatory regardless of extent. In fact, it can be speculated from our data that it is when hypertrophy is inadequate to normalize wall stress that the transition to pathological hypertrophy occurs.

View this table: [in this window] [in a new window]   Table 3. Extent of Hypertrophy and Its Effect on Contractile Function

Assessment of Contractile Function: Correlation of In Vivo and In Vitro Measurement
In this study, the presence or absence of contractile dysfunction was used as the key variable for dichotomizing the groups of dogs because it is the presence of muscle dysfunction that helps distinguish pathological from adaptive hypertrophy. The assessment of contractile function in vivo has been fraught with difficulty. The perfect index of contractility would be independent of load, independent of left ventricular mass and size, and sensitive to changes in contractile function. In previous studies we have used an inferior vena cava balloon to alter pressure and volume to construct various indexes of contractility.^{7 8} This technique was not used in the current studies because we feared that balloon-induced hypotension might cause ischemia in the pressure-overloaded myocardium known to be at risk for ischemia.^{30 31 32} Therefore, we used the relationship of mean normalized systolic ejection rate and wall stress to infer contractile function. Mean normalized systolic ejection rate is similar to mean velocity of circumferential fiber shortening except that it relies on volumes rather than dimensions. As such, this index could be expected to be dependent on contractile function and afterload but reasonably independent of preload.¹¹ Since ejection fraction is dimensionless, this index is size-independent, unlike maximal systolic elastance, which varies inversely with ventricular volume.³³ By plotting MNSER against afterload (wall stress), an index that is dimensionless and relatively preload independent that incorporates afterload into its expression is produced. We and others have used this paradigm successfully to assess contractile function in the past.^{10 29} Importantly, our in vivo findings were corroborated by indexes of cell contractile function measured by investigators in our group, who were blinded to our results. This collaboration gave us the opportunity to corroborate the in vivo measurement with another standard of contractile function-the function of the cardiocytes isolated from the left ventricle that had been evaluated in vivo. Such corroboration helps strengthen our own conclusions and helps lend credence to past studies of in vivo contractile function performed by others using similar indexes in which a second standard of contractile function was unavailable. These studies further suggest that when ventricular contractile dysfunction is present, it is at least in part a property of the cardiocytes comprising the chamber as opposed to other chamber factors such as geometry, collagen deposition, etc.

Model of Pressure Overload
The hypertrophy that develops in experimental overload must be viewed in the context of the model that produces it. Models of right ventricular pressure-overload hypertrophy and left ventricular pressure-overload hypertrophy have been developed in the adult and juvenile animals. In juvenile animals, a fixed band is placed around the pulmonary artery or aorta, and as the animal grows the pressure overload gradually worsens. These models have the advantage of producing gradually induced overload such as might occur in humans but have the disadvantage of producing the hypertrophy in juveniles in which hypertrophy might be different from that which occurs in adults.^{14 34} While it is important to study juvenile hypertrophy, which affects many children with congenital heart disease, adult models probably should be studied if it is the pathophysiology of the adult that is sought. However, most previous adult models have used an initial fixed constriction, which, if mild, produces only mild pressure overload or if severe may cause sudden pressure overload, potentially producing myocardial damage.¹³ This damage may introduce artifact into the experimental results. Acute overload is also at variance with most pressure overload in adults in which it is acquired gradually rather than suddenly. Our model is specifically one of gradually developing hypertrophy in the adult canine. It overcame previous problems with aortic rupture, a specific problem in dogs, and allowed longitudinal study of the animals.

We conclude that in this model of gradually applied left ventricular pressure overload in the adult dog that outcome is predicated to a large extent on baseline differences in mass and wall stress that occur naturally in normal dogs and not to differences in the rate or the amount of pressure overload. These differences may be genetically mediated. The dogs that developed extensive hypertrophy and maintained normal function had higher ventricular mass despite lower wall stress at baseline, suggesting they were already responding differently to their ambient load than the animals that developed less hypertrophy and muscle dysfunction.

	Acknowledgments

 
This research was supported in part by National Institutes of Health grant PO1-HL-48788 (George Cooper IV, Program Director).

	Appendix 1

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 
Derivation of Modified Midwall Mean Vcf
Assuming that left ventricular (LV) mass is constant throughout the cardiac cycle, using a cylindrical model,

(1)

where L is end-diastolic long axis, R is end-diastolic internal radius, and h is end-diastolic wall thickness.

At a specific time in the cardiac cycle,

(2)

where L', R', and h' are the dimensions at that specific time in the cardiac cycle.

(3)

LV wall thickness is divided into inner and outer shells with thicknesses of a and b, respectively; a+b=h, and at end diastole, a=b.

(4)

where a is the wall thickness of the inner shell of the LV wall.

(5)

At a specific time in the cardiac cycle,

(6)

where a' is the wall thickness of the inner shell of the LV wall at that specific time in the cardiac cycle.

Assuming that the volume of each shell remains constant throughout the cardiac cycle,

(7)

By dividing Equation 7 by Equation 3,

(8)

Since the value of R, h, a, and R' are known, a' can be calculated.

To calculate midwall mean Vcf,

(9)

where Dd is end-diastolic LV diameter, Ds is end-systolic LV diameter, and hes is end-systolic wall thickness.

(10)

(11)

(12)

(13)

Received July 15, 1996; revision received October 23, 1996; accepted November 18, 1996.

	References

Top Abstract Introduction Methods Results Discussion Appendix 1 References

 

1. Grossman W. Cardiac hypertrophy: useful adaptation or pathologic process? Am J Med. 1980;69:576-584. [Medline] [Order article via Infotrieve]
   
2. Cooper G IV, Kent RL, Uboh CE, Thompson EW, Marino TA. Hemodynamic versus adrenergic control of cat right ventricular hypertrophy. J Clin Invest. 1985;75:1403-1414.
   
3. Gunther S, Grossman W. Determinants of ventricular function in pressure-overload hypertrophy in man. Circulation. 1979;59:679-688. [Free Full Text]
   
4. Huber D, Grimm J, Koch R, Krayenbuehl HP. Determinants of ejection performance in aortic stenosis. Circulation. 1981;64:126-134. [Free Full Text]
   
5. Donner R, Carabello BA, Black I, Spann JF. Diminished wall stress in compensated aortic stenosis. Am J Cardiol. 1983;51:946-951. [Medline] [Order article via Infotrieve]
   
6. Carroll JD, Carroll EP, Feldman T, Ward DM, Lang RM, McGaughey D, Karp RB. Sex-associated differences in left ventricular function in aortic stenosis of the elderly. Circulation. 1992;86:1099-1107. [Abstract/Free Full Text]
   
7. Urabe Y, Mann DL, Kent RL, Nakano K, Tomanek RJ, Carabello BA, Cooper G IV. Cellular and ventricular contractile dysfunction in experimental canine mitral regurgitation. Circ Res. 1992;70:131-147. [Abstract/Free Full Text]
   
8. Tsutsui H, Spinale FG, Nagatsu M, Schmid PG, Ishihara K, DeFreyte G, Cooper G IV, Carabello BA. Effects of chronic ß-adrenergic blockade on the left ventricular and cardiocyte abnormalities of chronic canine mitral regurgitation. J Clin Invest. 1994;93:2639-2648.
   
9. Peterson KL, Skloven D, Ludbrook P, Uther JB, Ross J Jr. Comparison of isovolumic and ejection phase indices of myocardial performance in man. Circulation. 1974;49:1088-1101. [Abstract/Free Full Text]
   
10. Colan SD, Borow KM, Neumann A. Left ventricular end-systolic wall stress-velocity of fiber shortening relation: a load-independent index of myocardial contractility. J Am Coll Cardiol. 1984;4:715-724. [Abstract]
    
11. Nixon JV, Murray RG, Leonard PD, Mitchell JH, Blomqvist CG. Effect of large variations in preload on left ventricular performance characteristics in normal subjects. Circulation. 1982;65:698-703. [Free Full Text]
    
12. Shimizu G, Zile MR, Blaustein AS, Gaasch WH. Left ventricular chamber filling and midwall fiber lengthening in patients with left ventricular hypertrophy: overestimation of fiber velocities by conventional midwall measurements. Circulation. 1985;71:266-272. [Abstract/Free Full Text]
    
13. Bishop SP, Melsen LR. Myocardial necrosis, fibrosis, and DNA synthesis in experimental cardiac hypertrophy induced by sudden pressure overload. Circ Res. 1976;39:238-245. [Abstract/Free Full Text]
    
14. Fujii AM, Aoyagi T, Flanagan MF, Takahasi T, Bennett-Guerrero E, Colan SD, Izumo S. Response of the hypertrophied left ventricle to tachycardia: importance of maturation. Am J Physiol. 1993;264:H983-H993. [Abstract/Free Full Text]
    
15. Krahwinkel DJ Jr, Sawyer DC, Eyster GE, Bender G. Cardiopulmonary effects of fentanyl-droperidol, nitrous oxide, and atropine sulfate in dogs. Am J Vet Res. 1975;36:1211-1219. [Medline] [Order article via Infotrieve]
    
16. Rackley CE, Dodge HT, Coble YD Jr, Hay RE. A method for determining left ventricular mass in man. Circulation. 1964;29:666-671. [Abstract/Free Full Text]
    
17. Corin WJ, Swindle MM, Spann JF Jr, Nakano K, Frankis M, Biederman RW, Smith A, Taylor A, Carabello BA. Mechanism of decreased forward stroke volume in children and swine with ventricular septal defect and failure to thrive. J Clin Invest. 1988;82:544-551.
    
18. Tomita M, Spinale FG, Crawford FA, Zile MR. Changes in left ventricular volume, mass, and function during the development and regression of supraventricular tachycardia-induced cardiomyopathy: disparity between recovery of systolic versus diastolic function. Circulation. 1991;83:635-644. [Abstract/Free Full Text]
    
19. Mirsky I. Left ventricular stresses in the intact human heart. Biophys J. 1969;9:189-208.
    
20. Nakamura T. BMDP program for piecewise linear regression. Comput Methods Programs Biomed. 1986;23:53-55. [Medline] [Order article via Infotrieve]
    
21. Sasayama S, Franklin D, Ross J Jr. Hyperfunction with normal inotropic state of the hypertrophied left ventricle. Am J Physiol. 1977;232:H418-H425. [Abstract/Free Full Text]
    
22. Aoyagi T, Fujii AM, Flanagan MF, Arnold LW, Brathwaite KW, Colan SD, Mirsky I. Transition from compensated hypertrophy to intrinsic myocardial dysfunction during development of left ventricular pressure-overload hypertrophy in conscious sheep: systolic dysfunction precedes diastolic dysfunction. Circulation. 1993;88:2415-2425. [Abstract/Free Full Text]
    
23. Carabello BA, Mee R, Collins JJ Jr, Kloner RA, Leven D, Grossman W. Contractile function in chronic gradually developing subcoronary aortic stenosis. Am J Physiol. 1981;240:H80-H86.
    
24. Newman WH, Webb JG. Adaptation of the left ventricle to chronic pressure overload: response to inotropic drugs. Am J Physiol. 1980;238:H134-H143.
    
25. Wisenbaugh T, Allen P, Cooper G, Holzgrafe H, Beller G, Carabello BA. Contractile function, myosin ATPase activity and isozymes in the hypertrophied pig left ventricle following a chronic progressive pressure overload. Circ Res. 1983;53:332-341. [Abstract/Free Full Text]
    
26. Gaasch WH, Zile MR, Hoshino PK, Apstein CS, Blaustein AS. Stress-shortening relations and myocardial blood flow in compensated and failing canine hearts with pressure-overload hypertrophy. Circulation. 1980;79:872-883. [Abstract/Free Full Text]
    
27. Cooper G, Tomanek RJ, Airhart JC, Marcus ML. Chronic progressive overload of the cat right ventricle. Circ Res. 1981;48:488-497. [Abstract]
    
28. Spann JF Jr, Buccino RA, Sonnenblick EH, Braunwald E. Contractile state of cardiac muscle obtained from cats with experimentally produced ventricular hypertrophy and heart failure. Circ Res. 1967;21:341-354. [Abstract/Free Full Text]
    
29. Carabello BA, Zile MR, Tanaka R, Cooper G IV. Left ventricular hypertrophy due to volume overload versus pressure overload. Am J Physiol. 1992;263(Heart Circ Physiol. 32):H1137-H1144.
    
30. Fujii AM, Gelpi RJ, Mirsky I, Vatner SF. Systolic and diastolic dysfunction during atrial pacing in conscious dogs with left ventricular hypertrophy. Circ Res. 1988;62:462-470. [Abstract/Free Full Text]
    
31. Nakano K, Corin WJ, Spann JF Jr, Biederman RWW, Denslow S, Carabello BA. Abnormal subendocardial blood flow in pressure overload hypertrophy is associated with pacing-induced subendocardial dysfunction. Circ Res. 1989;65:1555-1564. [Abstract/Free Full Text]
    
32. Rembert JC, Kleinman LH, Fedor JM, Wechsler AS, Greenfield JC Jr. Myocardial blood flow distribution in concentric left ventricular hypertrophy. J Clin Invest. 1978;62:379-386.
    
33. Nakano K, Sugawara M, Ishihara K, Kanazawa S, Corin WJ, Denslow S, Biederman RWW, Carabello BA. Myocardial stiffness derived from end-systolic wall stress and the logarithm of the reciprocal of wall thickness: a contractility index independent of ventricular size. Circulation. 1990;82:1352-1361. [Abstract/Free Full Text]
    
34. Aoyagi T, Mirsky I, Flanagan MF, Currier JJ, Colan SD, Fujii AM. Myocardial function in immature and mature sheep with pressure-overload hypertrophy. Am J Physiol. 1992;262:H1036-H1048.[Abstract/Free Full Text]
    

This article has been cited by other articles:

				D. Scholz, P. McDermott, M. Garnovskaya, T. N. Gallien, S. Huettelmaier, C. DeRienzo, and G. Cooper IV Microtubule-associated protein-4 (MAP-4) inhibits microtubule-dependent distribution of mRNA in isolated neonatal cardiocytes Cardiovasc Res, August 1, 2006; 71(3): 506 - 516. [Abstract] [Full Text] [PDF]

				K. Berenji, M. H. Drazner, B. A. Rothermel, and J. A. Hill Does load-induced ventricular hypertrophy progress to systolic heart failure? Am J Physiol Heart Circ Physiol, July 1, 2005; 289(1): H8 - H16. [Abstract] [Full Text] [PDF]

				H. Yamaguchi, J. Yoshida, K. Yamamoto, Y. Sakata, T. Mano, N. Akehi, M. Hori, Y.-J. Lim, M. Mishima, and T. Masuyama Elevation of plasma brain natriuretic peptide is a hallmark of diastolic heart failure independent of ventricular hypertrophy J. Am. Coll. Cardiol., January 7, 2004; 43(1): 55 - 60. [Abstract] [Full Text] [PDF]

				E. Braunwald Congestive heart failure: a half century perspective Eur. Heart J., May 2, 2001; 22(10): 825 - 836. [PDF]

				M. Koide, M. Hamawaki, T. Narishige, H. Sato, S. Nemoto, G. DeFreyte, M. R. Zile, G. Cooper IV, and B. A. Carabello Microtubule Depolymerization Normalizes In Vivo Myocardial Contractile Function in Dogs With Pressure-Overload Left Ventricular Hypertrophy Circulation, August 29, 2000; 102(9): 1045 - 1052. [Abstract] [Full Text] [PDF]

				Y. Nagatomo, B. A. Carabello, M. L. Coker, P. J. McDermott, S. Nemoto, M. Hamawaki, and F. G. Spinale Differential effects of pressure or volume overload on myocardial MMP levels and inhibitory control Am J Physiol Heart Circ Physiol, January 1, 2000; 278(1): H151 - H161. [Abstract] [Full Text] [PDF]

				H. Laine, C. Katoh, M. Luotolahti, H. Yki-Jarvinen, I. Kantola, A. Jula, T. O. Takala, U. Ruotsalainen, H. Iida, M. Haaparanta, et al. Myocardial Oxygen Consumption Is Unchanged but Efficiency Is Reduced in Patients With Essential Hypertension and Left Ventricular Hypertrophy Circulation, December 14, 1999; 100(24): 2425 - 2430. [Abstract] [Full Text] [PDF]