Early Versus Delayed Angiotensin-Converting Enzyme Inhibition in Experimental Chronic Heart Failure
Effects on Survival, Hemodynamics, and Cardiovascular Remodeling
Background The efficacy of ACE inhibitors in congestive heart failure (CHF) might be affected by the pathophysiological status present at the onset of treatment. We compared in a rat model the effects of ACE inhibition (lisinopril, 10 mg·kg−1·d−1) initiated early (1 week) or late (3 months) after myocardial infarction (ie, at time points corresponding to moderate or severe CHF without or with established cardiac remodeling).
Methods and Results Survival was improved by early treatment at 3 months (from 76% to 95%) and by both early and delayed treatment at 9 months (placebo, 28%; early, 90%; delayed, 61%). Delayed treatment was initiated in a more severe pathophysiological context of CHF than early treatment, illustrated in untreated rats by higher left ventricular (LV) end-diastolic and central venous pressures and by increased LV weight and LV cavity circumference. After 9 months, early and delayed treatments reduced systolic, LV end-diastolic, and central venous pressures. Both treatments also similarly decreased LV weight, LV cavity circumference, and LV collagen density.
Conclusions In this rat model of CHF, early and delayed ACE inhibitor treatments both increase survival and exert similar beneficial effects on cardiac hemodynamics and remodeling. Although early treatment prevents the development of ventricular dysfunction and remodeling, delayed treatment is capable of reversing cardiac hypertrophy and remodeling, as well as ventricular dysfunction. Thus, ACE inhibitors exert marked beneficial effects even when treatment is initiated late into the evolution of heart failure (ie, at a time of established ventricular dysfunction and remodeling).
ACE inhibitors are widely used in the treatment of CHF. Indeed, there is no doubt that this class of drugs can increase both survival and quality of life for humans with CHF characterized by severe1 2 or moderate3 4 alterations in cardiac function. Such beneficial effects in terms of survival have been demonstrated in experimental CHF in rats.5 Furthermore, ACE inhibitors have been shown to prevent the development of CHF.6
The efficacy of ACE inhibitors probably results from a combination of their hemodynamic, biological, and, in the long term, structural effects. In humans and in experimental models of CHF, ACE inhibitors decrease cardiac preload and afterload,7 8 9 reduce sympathetic activity,10 and induce both cardiac11 and vascular12 remodeling, although the relative role of these different properties in their overall beneficial effect is not clear. Moreover, in the context of post-MI failure, it is not clear how long after MI or ventricular dysfunction treatment with ACE inhibitors should be initiated. Indeed, the relative effectiveness of early versus late treatment with ACE inhibitors has not been studied and cannot be easily assessed in humans. In experimental studies, treatments can usually be considered preventive because they are commonly initiated very early after MI. In this case, ACE inhibitors may interfere with the initial healing processes or with cardiac remodeling if treatments are administered before those processes are completed. However, the efficacy of ACE inhibitors administered late after infarction (ie, in the presence of established ventricular dysfunction and cardiac remodeling) has not been studied experimentally. In humans, long-term ACE inhibition prevents further progressive remodeling13 in patients with advanced but not yet symptomatic LV dilation, but the effect on survival has not been assessed in this context.
The objectives of the present study were to compare the effects on survival, cardiac hemodynamics, and cardiovascular remodeling of treatment with an ACE inhibitor (lisinopril) initiated either early after MI (ie, when ventricular dysfunction was still moderate and cardiac remodeling had not occurred) or late in the development of MI (ie, in a setting of severe ventricular dysfunction and established cardiac remodeling).
Animals and Treatment
MI was produced in 10-week-old male Wistar rats (Iffa-Credo, L'Arbresle, France) through left coronary artery ligation, as previously described.5 Briefly, rats were anesthetized with ether, and a left thoracotomy was performed. The heart was exteriorized, and the left coronary artery was ligated ≈2 mm from its origin with a Prolene 6-0 suture. In sham-operated animals, the suture was passed but not tied. The heart was then rapidly replaced in the chest, which was closed. With this method, mortality over the first 48 hours was ≈35% for the infarcted group and 2% for the sham-operated group.
All infarcted rats were allowed 7 days to recover before being assigned to the untreated group (MI control) or to one of the lisinopril-treated groups (10 mg·kg−1·d−1 in drinking water). Treatment was initiated either 7 days (early treatment) or 3 months (delayed treatment) after ligation. Rats were weighed and their water intake was measured weekly to allow adjustment of the drug concentration of the drinking water, whereas drug solution was made every 2 days. All rats received standard rat chow and water (with or without lisinopril) ad libitum and were maintained on a 12-hour light/dark cycle.
Subgroups of randomly selected animals were then killed at 7 days (D0), 3 months (D90), or 9 months (D270) after MI.
The effect of lisinopril on survival was assessed in four groups (sham, 15 rats; MI control, 29 rats; early treatment, 20 rats; delayed treatment, 18 rats) and was followed over a 9-month period. During the treatment period, cages were inspected daily for animals that had died. For each dead animal, the heart was removed, cleaned, and immersed in Bouin solution for subsequent determination of infarct size.
Hemodynamic Measurements in Conscious Rats
Systolic blood pressure (plethysmography) was determined in surviving rats at D0 and D15 and every month for 9 months.
Hemodynamic Assessment in Anesthetized Rats
At selected time points (D0, D90, and D270), rats were anesthetized with pentobarbital (50 mg·kg−1 IP). The right carotid artery and the right external jugular vein were cannulated with a micromanometer-tipped catheter (SPR 407, Millar Instruments) and advanced into the aorta and thoracic vena cava, respectively, for recording of arterial and central venous pressures. The aortic catheter was then advanced into the LV for recording of pressures and maximal rate of rise (dP/dtmax) of LV pressure. All tracings were recorded on a physiological recorder, and heart rate was obtained from an arterial pressure tracing.
After cardiac hemodynamic parameters were obtained, a femoral artery was carefully isolated under a dissection microscope and transferred to an arteriograph14 (Living Systems Instrumentation) filled with oxygenated (95% O2/5% CO2) physiological saline solution of the following composition (×10−3 mol/L): NaCl 119, NaCO3 24, KCl 4.7, KH2PO4 1.18, MgSO4·7H2O 1.17, CaCl2 1.6, and glucose 5.5. The vessels were cannulated, stretched to their in vivo length, and pressurized using a servo-controlled pump connected to a micromanometer. The intraluminal pressure was set to 30 mm Hg, which corresponds to the level of pressure for which the vasoconstrictor response to 100 mmol/L KCl is maximal (ie, “optimal” pressure). This level of initial tone was also used previously in similar rat mesenteric artery preparations.15 The arteriograph chamber was placed on the stage of an inverted microscope connected to a black-and-white video camera. The intraluminal diameter of pressurized arteries was then continuously measured with an on-line image analyzer (Living Systems Instrumentation) connected to a video camera.14
After equilibration of the vessels for 45 to 60 minutes at optimal intraluminal pressure, the inhibition of the vascular tissue renin-angiotensin system by lisinopril was assessed by the effect on the femoral artery of angiotensin I (10−6 mol/L) and angiotensin II (10−6 mol/L).
Immediately after dissection of the femoral artery, the heart was arrested in diastole by an injection of KCl, excised, cleaned, weighed, and immersed in Bouin fixative solution. Twenty-four hours later, the heart was cut perpendicular to the apex-to-base axis into three sections of similar thickness. These sections were dehydrated and embedded in paraffin, and 3-μm-thick slices were cut, which were stained with Sirius red and mounted onto glass slides.
For the determination of infarct size, the slices were placed under a video microscope (Microwatcher VS-30H, Mitsubishi Kasei Corp) with a ×20 enlargement lens. The camera was connected to an electronic color digitalization card (Matrox Illuminator 16), and the digitized color images were stored on an IBM-compatible computer. The stored images were later displayed on a 1024×768-pixel color screen using Windows-based image analysis software (Cyberview, Cervus International); the endocardial and epicardial circumferences of the infarcted portion and the entire LV were determined using the same software. Infarct size was expressed as the ratio of the sum of the endocardial and epicardial circumferences of the infarcted portion to the sum of the entire LV.5
For determination of cardiac collagen density, images (×500 enlargement) were obtained in the noninfarcted subepicardial region of the LV on Sirius red–stained slides. Volume collagen fraction was calculated as the sum of the surface of all connective tissue areas divided by the total surface of the image. Perivascular collagen was excluded from this measurement. It has been shown that total volume fraction, as determined with this morphometric approach, is closely related to hydroxyproline concentration of the ventricle.16 17
After the in vitro responses to angiotensin I and angiotensin II were obtained, the arteries were maximally dilated with nitroprusside and fixed at 30 mm Hg with Bouin fixative solution for 15 minutes, after which they were removed from the perfusion apparatus and immersed in Bouin solution. After fixation, arteries were dehydrated and embedded in paraffin. Three-micrometer-thick slices were cut perpendicular to the vessel axis and stained with either Sirius red (for the determination of collagen) or Orcein (for the determination of elastin). Slides were enlarged ×1500 to determine media collagen and elastin density by image analysis. Cross-sectional area of the media, limited by the internal and external lamina, was also determined in Orcein-stained slides at a magnification of ×500.
All reported values, except survival, are given as mean±SEM. Comparison of survival in untreated and lisinopril-treated MI rats was performed using the Mantel procedure.18
Differences between values obtained 1 week after the surgical procedure in sham-operated and untreated MI groups were evaluated by unpaired t test. Differences among values obtained at D0, D90, and D270 were evaluated with an ANOVA for each time interval. If the ANOVA revealed significance, it was followed by a Tukey test for multiple comparisons. Differences were considered significant at the level of P<.05.
During the 9-month observation period, none of the sham-operated animals died. Fig ⇓1 illustrates the survival curves in the untreated CHF rats and in both lisinopril-treated groups.
Early Lisinopril Treatment
The survival curves of the untreated and the early lisinopril treatment groups diverged immediately after the onset of treatment; after 3 months, survival was higher in lisinopril-treated rats than in control animals (95% versus 76%; P<.05). After 9 months, survival was markedly and significantly improved by the early treatment (90% versus 28%, P<.05).
Delayed Lisinopril Treatment
Before the start of treatment (ie, during the first 3 months after MI), the mortality rates were comparable in the untreated and delayed lisinopril-treated groups. Indeed, after 3 months, survival was 76% and 78% in the untreated and delayed lisinopril-treated groups, respectively. Delayed treatment with lisinopril improved survival at the end of the observation period (61% versus 28%; P<.05).
The marked effect of lisinopril on survival was not due to differences in infarct size in the three groups because mean infarct size, including both spontaneously deceased and surviving animals, was 54±4%, 50±3%, and 51±3% in untreated, early treatment, and delayed treatment groups, respectively (P=NS).
Hemodynamic Measurements in Conscious Rats
Fig ⇓2 shows the evolution of systolic blood pressure in the surviving rats from the four groups during the 9-month study period. Systolic blood pressure of untreated CHF rats was always significantly lower than that of sham-operated animals. Moreover, although systolic blood pressure levels of control and both lisinopril-treated CHF groups were similar before the start of the treatment (101±3, 101±2, and 99±2 mm Hg, respectively), early treatment induced a reduction in systolic blood pressure compared with control rats, which was already significant after 2 weeks (control, 110±2 mm Hg; lisinopril, 88±2 mm Hg; P<.05), and this effect persisted throughout the study. Indeed, after 9 months of treatment, systolic blood pressure was 132±2 and 95±3 mm Hg in untreated and lisinopril-treated animals, respectively (P<.05). Regarding the delayed treatment, during the first 3 months of the study (ie, before treatment), systolic blood pressure was identical to that of untreated animals. After 1 month of treatment (ie, 4 months after MI), systolic blood pressure was significantly reduced compared with that of untreated animals (94±4 versus 118±4 mm Hg, P<.05). This decrease was approximately the same magnitude as that induced by early treatment (86±2 mm Hg). This effect on blood pressure remained stable throughout the treatment period.
Hemodynamic Measurements in Anesthetized Rats
Fig ⇓3 illustrates the cardiac hemodynamics and central venous pressure measured in anesthetized animals before treatment and after 3 or 9 months of treatment. Heart rate was never affected by CHF or its treatment with lisinopril (data not shown). Compared with sham-operated animals, CHF decreased LV systolic pressure and LV dP/dtmax, and these effects were significant at the three time points studied. LV systolic pressure was significantly reduced after 3 months by the early lisinopril treatment. After 9 months, early as well as delayed treatment reduced LV systolic pressure. In contrast, LV dP/dtmax was never significantly affected by the treatment.
Compared with sham-operated animals, CHF increased LVEDP and central venous pressure. Both pressures were normalized at D90 by early lisinopril treatment and at D270 by early as well as by delayed treatment.
Fig ⇓4 illustrates infarct size, heart weight, LV cavity circumference, and LV collagen density. Neither time nor lisinopril treatment affected the infarct size of animals killed at D0, D90, and D270. In sham-operated animals, heart weight and LV collagen density increased with time, whereas LV cavity circumference remained stable. Compared with time-matched sham-operated animals, CHF induced significant increases in heart weight, LV cavity circumference, and LV collagen density. Compared with time-matched CHF control animals, early treatment with lisinopril decreased heart weight at D90. At D270, both early and delayed treatments reduced heart weight to the same extent.
LV cavity circumference was normalized by early treatment at D90 but only slightly affected at D270. Similarly, at D270, LV circumferences were only marginally reduced by delayed treatment. LV collagen density was significantly reduced by early treatment at D90 and D270, as well as by delayed lisinopril at D270.
Fig ⇓5 shows internal diameter, media cross-sectional area, and collagen density measured in the femoral artery at the three time points. In sham-operated animals, media cross-sectional area and collagen density increased with time. Compared with sham-operated animals, CHF did not induce any changes in media cross-sectional area or collagen density and did not affect vascular elastin density (data not shown). Compared with time-matched CHF control animals, early lisinopril treatment decreased media cross-sectional area and collagen density, and this was significant at both D90 and D270. At D270, delayed lisinopril treatment induced the same reduction in cross-sectional area and collagen density as did early treatment.
The responses of isolated pressurized femoral arteries to angiotensins I and II (10−6 mol/L) are shown in the Table⇓. In arteries from sham-operated or untreated CHF rats, the responses to angiotensin I or angiotensin II were similar and were not affected by time. Early lisinopril treatment significantly reduced the responses to angiotensin I (at D90 for early treatment and at D270 for both early and delayed treatments) but did not affect the response to angiotensin II.
Results of the present study, which was performed in a rat model of CHF, show that both early (7 days after MI) and delayed (3 months after MI) treatments with an ACE inhibitor increase survival and exert beneficial effects on cardiac hemodynamics and remodeling. Moreover, at 9 months after MI, the beneficial effects of the early and late treatments on cardiac hemodynamics (as assessed on the basis of decreased LVEDP and central venous pressure) and on cardiac remodeling (as assessed on the basis of the effect on hypertrophy, ventricular dilatation, and ventricular collagen accumulation) appear to be quantitatively similar. Thus, given the fact that the early treatment was initiated in a setting of moderate ventricular dysfunction and cardiac remodeling, whereas late treatment was initiated in a context of severe LV dysfunction and established remodeling, our experiments suggest that ACE inhibitors are capable of both preventing (in the case of early treatment) and reversing (in the case of late treatment) the development of cardiac dysfunction and remodeling.
The beneficial effect of chronic ACE inhibition on survival was first demonstrated in a rat model of MI similar to the one used in the present study.5 Subsequent clinical trials, with survival as a major end point, have extended these results to humans with chronic CHF.1 2 3 Thus, the rat model has proved to be extremely useful in the study of the progression from LV dysfunction to overt CHF, as well as in the investigation of the beneficial effects of ACE inhibitors.8 19 20 Thus, our results obtained with the early treatment confirm previous findings obtained with other ACE inhibitors in the rat model. However, to the best of our knowledge, this is the first study to assess the effect of a delayed ACE inhibitor treatment on survival, as well as the long-term effect of such a delayed treatment of cardiac function and structure.
The use of early versus delayed treatment allowed us to study the effect of an ACE inhibitor administered in two different pathophysiological contexts. Indeed, at the time of initiation of the early treatment (ie, 7 days after MI), the decrease in arterial blood pressure and LV dP/dt and increase in central venous pressure and LVEDP indicate the presence of moderate LV dysfunction. Similarly, at this time, LV remodeling is moderate, as illustrated by the slight increase in LV weight and cavity circumference. However, the hemodynamic context significantly worsens with time, since central venous pressure and LVEDP markedly increase between 1 week and 3 months. Indeed, after 3 months (ie, when we initiated the delayed treatment), these hemodynamic parameters are markedly increased not only versus time-matched sham-operated animals but also versus the CHF group as assessed at 7 days. Furthermore, at this time, the hemodynamic deterioration is associated with profound cardiac remodeling and collagen accumulation. Thus, in this study, early lisinopril therapy has been initiated in a less severe context of CHF than delayed treatment.
After long-term administration, both early and delayed lisinopril therapy8 12 reduced cardiac hypertrophy, LV remodeling, and interstitial collagen density in the noninfarcted LV. Normalization of cardiac preload and reduction in afterload, which are probably associated with a decrease in ventricular wall stress, together with the neurohumoral effects of ACE inhibition on myocyte growth and collagen synthesis are probably responsible for the structural effects of ACE inhibitors on the cardiovascular system. Nevertheless, it should be stressed that in contrast with delayed lisinopril therapy, profound cardiac remodeling is “absent” when early lisinopril therapy is initiated. Thus, early treatment seems to slow down the progression of cardiac remodeling, whereas delayed therapy regresses this remodeling. Indeed, early treatment reduces or normalizes LV hypertrophy, dilation, and fibrosis, not only at the end of the survival study but also after 90 days of treatment.
In our experiments, chronic CHF did not affect media cross-sectional area and media collagen or elastin densities, which is in agreement with previous results.12 21 Despite this lack of effect of CHF, ACE inhibition induces a decrease in media cross-sectional area and collagen density, without a modification in elastin content. This could be the result of the decrease in mean arterial pressure (ie, wall stress) and/or of the direct tissue effects of ACE inhibitors on the vascular wall mediated by the changes in angiotensin II and bradykinin. It could be assumed that structural vascular changes induced by ACE inhibitors improve blood flow delivery and regulation. In myocardial tissue, for example in addition to the reduction in LV hypertrophy, dilation, interstitial collagen accumulation, and wall stress, the vascular effects of ACE inhibition could contribute to improvement of the balance between myocardial oxygen demand and delivery and diminish arrhythmogenicity, which results in reduced mortality. Again, however, these vascular effects appear to be similar in the two treatment groups, suggesting that these potential beneficial effects can be obtained with both early and delayed treatment.
In conclusion, this study performed in a rat model of CHF shows that delayed treatment with an ACE inhibitor increases survival and exerts beneficial effects on cardiac hemodynamics and remodeling that are similar to those induced by an early treatment. Given the time course of the functional and structural cardiovascular changes observed in the present study, this suggests that although early treatment prevents the development of ventricular dysfunction and remodeling, delayed treatment is capable of reversing cardiac hypertrophy and remodeling, as well as ventricular dysfunction. These findings suggest that ACE inhibitors exert marked beneficial effects even when initiated late into the evolution of CHF (ie, at a time of established ventricular dysfunction and remodeling).
Selected Abbreviations and Acronyms
|CHF||=||congestive heart failure|
|LV||=||left ventricular, left ventricle|
|LVEDP||=||left ventricular end-diastolic pressure|
This study was supported in part by Zeneca, Paris, France.
- Received September 9, 1996.
- Revision received October 17, 1996.
- Accepted October 23, 1996.
- Copyright © 1997 by American Heart Association
Pfeffer MA, Pfeffer JM, Steinberg C, Finn P. Survival after an experimental myocardial infarction: beneficial effects of long-term therapy with captopril. Circ Res. 1985;72:406-412.
Wollert KC, Studer R, Just H, Drexler H. Influence of lisinopril on long-term mortality in rats with chronic myocardial infarction: role of blood pressure reduction and tissue ACE inhibition. Circulation. 1994;90:2457-2467.
Hirsch AT, Talsness CE, Schunkert H, Paul M, Dzau VJ. Tissue specific activation of the cardiac angiotensin converting enzyme in experimental heart failure. Circ Res. 1991;69:475-482.
Michel JB, Heudes D, Michel O, Poitevin P, Phillippe M, Scalbert E, Corman B, Levy B. Effect of chronic ANG I-converting enzyme inhibition on aging processes, II: large arteries. Am J Physiol. 1994;267:R124-R135.
Mulder P, Elfertak L, Richard V, Compagnon P, Devaux B, Henry JP, Scalbert E, Desche´ P, Mace´ B, Thuillez C. Peripheral artery structure and endothelial function in heart failure: effect of ACE inhibition. Am J Physiol. 1996;271:H469-H477.
Gaudron P, Kugler I, Hu K, Eilles C, Ertl G. Quinapril, initiated after advanced ventricular dilation in patients with chronic infarction, prevents progressive remodeling. J Am Coll Cardiol. 1996;27:228A. Abstract.
Dohi Y, Thiel MA, Bu¨hler FR, Lu¨scher TF. Activation of endothelial l-arginine pathway in resistance arteries: effect of age and hypertension. Hypertension. 1990;15:170-179.
Weber KT, Janicki JS, Schroff SG, Pick R, Chen RM, Bashey RI. Collagen remodeling of the pressure-overloaded, hypertrophied nonhuman primate myocardium. Circ Res. 1988;62:757-765.
Nicoletti A, Heudes D, Hinglais N, Appay M-D, Philippe M, Sassy-Prigent C, Bariety J, Michel J-B. Left ventricular fibrosis in renovascular hypertensive rats: effect of losartan and spironolactone. Hypertension. 1995;26:101-111.
Pfeffer JM, Pfeffer MA, Braunwald E. Influence of chronic captopril therapy on the infarcted left ventricle of the rat. Circ Res. 1985;57:84-95.
Heeneman S, Leenders PJA, Aarts PJJW, Smits JFM, Arends JW, Daemen MJAP. Peripheral vascular alterations during experimental heart failure in the rat: do they exist? Arterioscler Thromb Vasc Biol. 1995;15:1503-1511.