Cardiovascular Effects of Insulin-Like Growth Factor-1 and Growth Hormone in Chronic Left Ventricular Failure in the Rat
Background Insulin-like growth factor-1 (IGF-1) appears to have favorable cardiac effects associated with left ventricular remodeling early after myocardial infarction in the rat. The present study was designed to determine whether IGF-1 combined with growth hormone would be beneficial later as well, when infarct healing and cardiac remodeling have occurred.
Methods and Results Four weeks after coronary occlusion, 36 rats were randomized to IGF-1 (3 mg·kg−1·d−1) plus growth hormone (0.1 mg BID) or to placebo for 4 weeks. Treated rats had significant increases in body weight (22%), while the ratio of heart weight to body weight was unchanged. Under anesthesia, cardiac output (fluorescent microspheres) increased 46%, and systemic vascular resistance decreased by 21% (P<.001) in the treated group; a significant (22%) increase of the cardiac index was limited to treated rats with large myocardial infarctions. Small increases in the reduced left ventricular ejection fractions and left ventricular dP/dtmax values with treatment were not significant. Treated rats showed a borderline (16%) increase in left ventricular end-diastolic volume (angiography), whereas the ratio of left ventricular end-diastolic volume to body weight was reduced in the treated group.
Conclusions IGF-1 plus growth hormone administered to rats with left ventricular failure starting 1 month after MI was associated with substantial body growth, decreased systemic vascular resistance, and increased cardiac output. The failing heart also underwent treatment-induced increases in left and right ventricular weights in proportion to body growth, but left ventricular remodeling was minor, and a decrease in the ratio of left ventricular end-diastolic volume to body weight reflected relatively less chamber dilation compared with controls. A significant interaction between size of the myocardial infarction and treatment was observed for several variables, and IGF-1 and growth hormone increased the cardiac index (P<.035) in rats with a large myocardial infarction.
Marked hypertrophy occurs in viable regions of myocardium in rats with a large MI, yet severe LV dysfunction later ensues.1 2 3 4 Previous studies5 6 7 8 9 on LV remodeling after acute MI focused primarily on the interplay between infarct expansion, ventricular dilation, and compensatory hypertrophy in determining LV function. Whether such compensatory hypertrophy is beneficial or maladaptive in heart failure has remained unsettled, particularly in view of studies that show beneficial effects in rats10 and human subjects11 12 with use of ACE inhibitors, which also cause inhibition of cardiac hypertrophy.13 On the other hand, therapy to enhance the degree of compensatory hypertrophy after MI might also prove beneficial. We recently showed14 in normal rats that administration of IGF-1 for 2 weeks stimulates cardiac hypertrophy, and when given for 2 weeks starting 2 days after acute MI, it influences ventricular remodeling. Thus, compared with untreated rats with infarction, IGF-1–treated animals showed increases in LV weight, LVEDV, and estimated stroke volume, although body weight also was augmented and the ratios of these variables to body weight were unchanged. LV end-diastolic pressures were not affected by treatment, and the LV ejection fraction increased in IGF-1–treated rats with large infarctions.14 Thus, the short-term effects of growth factor administration appeared to be beneficial rather than maladaptive in developing heart failure.
IGF-1, a 70–amino acid basic peptide, is an essential growth factor for cellular proliferation and differentiation during development.15 In addition, neonatal rat myocytes synthesize IGF-1 and express IGF-1 receptors, which suggests that IGF-1 can act in an autocrine or paracrine manner to modulate myocyte growth and hypertrophy in the developing heart.16 The addition of IGF-1 to cultured neonatal cardiomyocytes induces myocyte hypertrophy with increased expression of mRNA for muscle-specific genes (MLC-2 and troponin-1),17 and when added to adult rat cardiomyocytes, it stimulates cardiac protein synthesis.18 IGF-1 mRNA levels and protein content are increased in the left ventricle 3 to 5 weeks after induction of hypertension in the rat,19 20 which suggests that IGF-1 may be a physiological mediator of hypertrophy.
GH, a 191–amino acid protein produced by the anterior pituitary gland, causes cardiac enlargement,21 although it is unclear to what degree these effects are mediated by IGF-1. There are conflicting data as to whether or not GH injections22 or infusions23 stimulate the local expression of IGF-1 mRNA in rat cardiac muscle. However, GH administration increases systemic IGF-1 production and raises IGF-1 blood levels, which provides an alternative mechanism whereby IGF-1 may indirectly mediate some of the cardiac effects of GH. Therefore, it is unclear whether or not IGF-1 and GH have differential and independent effects on the heart, as they do on some other tissues.24 25
Administration of IGF-1 can increase the IGFBP levels in the blood, although this effect is insufficient to maintain adequate IGF-1 binding capacity.26 However, combined treatment with GH and IGF-1 in human subjects increased the concentration of IGFBP III, which led to higher serum concentrations of IGF-1 than IGF-1 administration alone.27 In addition, the combination of IGF-I and GH produced greater levels of nitrogen retention and attenuated the hypoglycemia induced by IGF-I alone.28
In our prior study,14 IGF-1 showed beneficial effects when it was begun early (at 2 days) after MI, before infarct healing at the onset of LV remodeling, and was continued for 2 weeks, although the cardiac output was not directly measured. Therefore, the present study was designed to determine whether or not more prolonged treatment with IGF-1 and GH would beneficially affect the cardiac output, cardiac size, and function later, when MI healing was complete and substantial LV remodeling had occurred. Also, an influence of MI size on therapeutic responses, detected in our initial study,14 was further assessed in the present investigation.
Animals were handled according to the animal-welfare regulations of the University of California at San Diego, and the experimental protocol was approved by its Animal Subjects Committee.
Female Sprague-Dawley rats (weight, 250 to 300 g) were anesthetized with a mixture of ketamine hydrochloride (100 mg/kg), xylazine (10 mg/kg), and morphine sulfate (5 mg/kg) given intraperitoneally. Complete occlusion of the left coronary artery was performed as described previously.4 14 In brief, anesthetized animals were ventilated under positive pressure, a left thoracotomy was performed, the pericardium was opened, and the left coronary artery, which is intramural, was encircled with a curved needle and 6-0 silk suture. After the ligature was tied, complete occlusion was evidenced by a regional color change of the myocardium, together with the appearance of acute ST-segment elevations on the ECG. The chest was then closed in layers and the pneumothorax evacuated.
After the operation, the animals were caged in proportion to their size, given water and standard rat chow ad libitum, and housed in a climate-controlled environment subjected to 12-hour light/dark cycles. Of the 52 animals subjected to operation and surviving to the 28th postoperative day, 36 (69%) showed clear ECG and echocardiographic evidence of infarction and were assigned randomly to treatment with placebo or the combination of IGF-1 and GH.
The echocardiographic evaluation was performed before randomization in all rats. Animals were anesthetized in the manner described above, the chest was shaved, and transthoracic ultrasonographic evaluation was performed with a 5-MHz transducer and a two-dimensional echo color Doppler system (HP 1000, Hewlett Packard). Parasternal long-axis and apical two-chamber views were obtained with the rat in a supine position. Evidence of MI was provided by the presence of a wall-motion abnormality. Color pulse-wave Doppler analysis was used to detect the presence or absence of a systolic jet into the left atrium characteristic of mitral regurgitation; no evidence of mitral regurgitation was identified in any of the animals.
The anesthetized animals were then placed in the supine position, a 1-cm midline incision was made in the abdominal wall, and an osmotic pump (model 2ML4, Alzet) was implanted in the peritoneal cavity. In the treatment group, pumps were filled with rhIGF-1 diluted in sodium acetate buffer to deliver a dose of 3 mg·kg−1·d−1 (body weight at implantation) for 4 weeks. In the placebo group, pumps were filled with sodium acetate buffer. The peritoneal cavity was then closed in layers. Treated animals also received 0.1 mg of recombinant hGH in 0.25 mL of sterile water, and control animals received 0.25 mL of normal saline by subcutaneous injection twice daily for 28 days.
Twenty-eight days after pump implantation, rats were anesthetized again with ketamine and xylazine and were weighed, and a 1-cm incision was made in the abdomen. The osmotic pump was removed, and any fluid present in the abdomen was drained. Six of the treated animals and one control animal had ascitic fluid (the fluid was clear; average of 72 mL in treated animals, 72 mL in the control). Animals were then reweighed, and the latter weight is reported as the final body weight. The rats with ascitic fluid had a higher LV ejection fraction and smaller infarct circumference, but there were no other significant differences from controls, although the IGF-1 levels tended to be slightly higher. Therefore, the presence of abdominal fluid was not related to the degree of heart failure but might have been due to a local reaction to the osmotic pump or the infusate.
Systemic Arterial Pressure in the Conscious State
Measurements were made in the awake state by use of a tail-cuff pressure monitor (IITC Life Science USA).14 Conscious animals were familiarized with a plastic restraining cage before measurements were obtained. Heart rate and blood pressures were measured immediately before randomization and at 2 and 4 weeks of treatment. A minimum of five measurements was obtained in each animal at the three time points, and the results were averaged.
Aortic and LV Pressures in the Anesthetized State
Subsequent to 4 weeks of treatment, animals were anesthetized as described above and placed in the supine position. Under a dissecting microscope, the right carotid artery was cannulated and a PE 50 tubing advanced into the ascending aorta, where arterial pressure was recorded. A 0.014-in guidewire was then inserted into the PE 50 tubing and advanced across the aortic valve under fluoroscopic guidance. The tubing was then advanced over the wire into the left ventricle, the guidewire was withdrawn, and LV pressure and its maximum first derivative (LV dP/dtmax) were measured. Recordings were made at a sampling rate of 1000/s with Codas software (Dataq Instruments, Inc). It is recognized that a fluid-filled catheter system does not have a sufficient frequency response to precisely measure absolute LV dP/dtmax, and therefore this determination was used only as an index of this variable for comparison of directional changes in the two groups.29 Hemodynamic tracings were analyzed by beat averaging of 10 consecutive beats with Cordat software.14 MAP was calculated as the diastolic pressure plus one third of the pulse pressure. SVR was calculated as (MAP−RAP)/CO (where CO is cardiac output and RAP is the mean right atrial pressure) and expressed as dynes-s/cm−5.
With the catheter in the left ventricle, an incision was made over the right groin, and a PE 50 tubing was advanced via the femoral artery into the right common iliac artery. Cardiac output was measured by a technique previously described in rats30 but with 15-μm fluorescent-labeled microspheres used instead of radioactive microspheres.31 The microspheres (9.77×106) were suspended in 10 mL of 0.1% Tween 80 and 0.9% saline (Molecular Probes, Inc), and vials were vortexed for a minimum of 10 minutes before injection to ensure adequate mixing. Reference blood sampling (0.95 mL/min) via the femoral artery catheter was begun 15 to 20 seconds before injection and continued for a minimum of 30 seconds after the LV catheter was flushed. Approximately 2.9×104 spheres (0.3 mL of the suspension) were injected into the left ventricle over a 15-second period with a Hamilton injection syringe (model 750), and the syringe and LV catheter were flushed with 1 mL normal saline for an additional 20 to 30 seconds. The reference sample volume was equivalent to the volume of the microsphere suspension and saline injected (1.3 mL).
Measurements of cardiac output were performed in triplicate in each animal. Six normal animals were studied in addition to those with MIs. A single colored microsphere was used for each animal, and a series of four to six animals was studied on the same day. The number of microspheres injected was determined by measurement of the total fluorescence in 0.3 mL of spheres obtained from the same Hamilton syringe, stopcock, and catheter with 1.0 mL saline flush. Four to six such injections were performed in separate test tubes from each 10-mL vial of microspheres. The mean fluorescence obtained from the four to six injections was used to measure the quantity of spheres (fluorescence) injected in the calculation of cardiac output for the series of animals studied on that day. The coefficient of variation for fluorescence from 112 such microsphere injections was 1.2% (range, 0.4% to 2.3%). Cardiac output (mL/min) was calculated as follows:
Each milliliter of the reference blood sample was digested in 0.8 mL of 16 mol/L KOH plus 1.4 mL of 2% Tween 20 in a 45°C water bath for 48 hours. After digestion, samples were filtered through 10-μm polycarbonate membrane filters (Poretics Corp, No. 11078) mounted in syringe filter holders (Poretics Corp, No. 91130) on a vacuum manifold. Filters were then dried and the fluorescent dyes extracted with 2 mL cellosolve acetate (Fischer Scientific, Inc, No. E181-4). The amount of fluorescent dye present in each sample was measured on a Perkin-Elmer LS50B luminescence spectrometer. The photomultiplier tube voltage was set at 820 V, and the excitation and emission slit widths were set at 4 and 5 nm, respectively. A cutoff filter eliminated all light below 350-nm wavelength.
After measurement of cardiac output, LV angiography was performed as previously described.4 14 Briefly, animals were intubated and mechanically ventilated, and a dissecting microscope was used to expose the right external jugular vein and cannulate it with PE 50 tubing, which was advanced into the superior vena cava. Nonionic contrast medium (0.5 mL, Omnipaque 350, Winthrop) was injected into the superior vena cava over a period of 2 seconds with a power injector (Cordis Corp) during brief suspension of respiration. When necessary, additional doses of anesthesia (ketamine 100 mg/kg, xylazine 10 mg/kg, or morphine 2.5 mg/kg IP) were administered to suppress spontaneous respiration during angiography. Angiograms were obtained by use of the constant fluoroscopic technique (60 kVP and 1 to 2 mA), with a General Electric Fluoricon 300 (model MSI-1250 III). The data were acquired in a 2.25-in field of view of the image intensifier and recorded on 1-in videotape by use of a standard interlaced scanning mode, with a 1-cm grid and a lead marker placed at the level of the heart, as described elsewhere.4 Each animal was studied sequentially in 30° right anterior oblique and 60° left anterior oblique projections with separate contrast injections.
Image Acquisition and Analysis
The details of the image processing and validation of analytic methods have been described.4 In brief, x-ray images were digitized at 30 frames/s with a resolution matrix of 512×512 pixels with 256 shades of gray by use of a time-base corrector and a video-processing system (GW Hannaway and Associates) interfaced to a computer system (Silicon Graphics model 4D/380VGX). Densitometric analysis of a digitized loop was used to identify end diastole and end systole. LVEDV and LVESV, stroke volume, and LV ejection fraction were calculated for two consecutive beats by use of the biplane area-length method,32 and the data were averaged.4
Postmortem and Histological Preparations
Diastolic Pressure-Volume Relations
To assess the passive pressure-volume characteristics of the left ventricle, LV balloons were fashioned by stretching a polypropylene membrane into the approximate shape of the LV chamber and attaching it to a plastic tube. The pressure-volume characteristics of the balloons were assessed before use to ensure that they exhibited a flat pressure response to volume infusion over the range studied in that heart in vivo, which would indicate that the pressure measured during a study was due only to the compliance of the left ventricle. After euthanasia, hearts were arrested with KCl solution and excised. The left atrial appendage was opened to permit balloon insertion through the mitral valve into the LV. The atrioventricular groove was then ligated around the tubing. A Hamilton microliter syringe (model 750) was used to inject 0.3- to 0.5-mL aliquots of saline while ventricular pressure was measured. Reproducible pressure-volume curves were generated over a pressure range of 0 to 40 mm Hg within 10 minutes of cardiac arrest.33 The volume of the empty balloon was measured by displacement and added to the measured volume to determine true volume of the left ventricle. Curves were fitted by use of a monoexponential relation, and the pressure-volume relations were also plotted in a semilog format and the slopes determined as a measure of LV chamber compliance.
Cardiac fixation was performed as previously described.9 14 Briefly, polyethylene catheters were introduced into the LV apex and the aortic root, and the LV chamber was filled from a reservoir and maintained at 10 mm Hg. After washout of blood from the coronary circulation with heparinized saline (10 000 U/L) for 3 minutes, the myocardium was perfused retrograde from the aorta with 10% formalin at a constant pressure of 60 mm Hg for 10 minutes. The right atrium and pulmonary artery were opened to decompress the right ventricle during fixation. The heart was then excised and immersed in 10% formalin for 24 hours. Subsequently, the atria and adhesions were dissected away, and the right and left ventricles were separated and weighed (the interventricular septum was included with the left ventricle). To compare heart weights among groups, chamber weights were normalized by the length of the tibia as well as by body weight. The right tibia was dissected, and its length from the condyles to the tip of the medial malleolus was measured with a micrometer caliper by the method of Yin et al.34
Determination of MI Size
The LV was embedded in paraffin, and serial transverse sections 10 μm thick were cut every 100th section from apex to base (1 mm apart), mounted, and stained with Milligan’s trichrome. Slides taken at the fifth and sixth millimeter from the apex were analyzed blindly to assess infarct size.4 14 35 The slides were projected with a microprojector (Jena) at a magnification of 13, and measurements were made by computerized planimetry. The percent infarct size, estimated from the ratio of the area of the infarct zone to the total myocardial area (infarcted and noninfarcted regions), previously was shown to correlate well with total infarct size calculated from all serial sections.35 Late after infarction, there is a loss of tissue volume reflected in wall thinning in the infarcted zone such that determination of infarct size by area relative to total LV area may lead to an underestimation of infarct size; therefore, the percent infarct size based on the length of the scar was estimated from the ratio of the sum of the scar lengths along the endocardial and epicardial surfaces to the sum of the total endocardial and epicardial circumferences.10 36 As with infarct size based on scar area, the fifth and sixth slices from the apex were used. Thickness of the scar was measured at several points along the circumference of the fifth and sixth slices, the center of the left ventricle was identified, and 12 cords were extended from the center to the endocardial surface. At each point where the cord intersected the endocardium, the shortest distance across the endocardial wall was measured (junctions between scar and noninfarcted myocardium were excluded). Thickness measurements of the scar were averaged for each animal and expressed in millimeters.
Before euthanasia, 1 mL of blood was obtained from the inferior vena cava, and the plasma was obtained and stored at −20°C. GH was not given on the day of the experiment, and the osmotic pump that delivered IGF-1 was not removed until after death. The blood was centrifuged, IGF-1 was separated from binding proteins by acid/ethanol precipitation, and total IGF-1 levels were measured by RIA. The RIA used rhIGF-1 as standard and [125I]IGF-1 as the label, and a polyclonal rabbit antibody to IGF-1 was added to measure bound and free total IGF-137 ; this method yields 91% recovery of IGF-1.38 hGH also was measured.39
Anti–IGF-1 antibodies also were determined to assess the possibility that antibodies to rhIGF-1 were induced during the 4-week treatment period. IGF-1 antibodies in rat serum were measured by an ELISA with positive and negative control sera used as reference standards. Microtiter plates were coated with 0.1 mL of 2 μg/mL IGF-1 for 16 hours at 4°C and blocked with PBS/0.5% BSA/0.05% polysorbate 20/0.01% thimerosal and 0.1 mL diluted rat-serum controls, reference standard was added, and the plates were incubated at room temperature for 2 hours. After washing, 0.1 mL of protein-A-HRP (Boehringer Mannheim, lot 605 295) was added, incubated for an additional 2 hours, and washed. Next, 0.1 mL of 0.2 mg/mL o-phenylenediamine and 0.01% H2O2 in PBS were added, then incubated in the dark for 25 minutes. The reaction was stopped by adding 0.1 mL of 4.5N H2SO4, and the plates were read with 490- to 492-nm filters used for absorbance and 405-nm filters used for reference.
All variables to be compared between treatment groups were correlated with infarct size because infarct size was expected to influence many of the measured variables and the possibility existed that the effects of IGF-1 and GH could be different depending on whether the infarct was small or large.14 Variables that showed no significant correlation with infarct size (P>.10, Pearson’s correlation), which included variables in Tables 1⇓ and 3⇓, were analyzed for differences between treated and control groups at 28 days with a Student’s t test (two tailed). Variables that were significantly correlated with infarct size were subjected to a two-way ANOVA, with treatment as one factor and infarct size less than or greater than the overall median (39.02%, expressed as percent of LV circumference) as the other. The probability value for the main effect (treatment) is reported, and if the interaction of treatment and infarct size was significant, the treatment difference within the large- and small-infarct groups was examined by a modified Student’s t test with the Bonferroni method used to adjust for multiple comparisons.
Among the 36 rats randomized to the experimental protocol (18 control, 18 treated), 2 treated animals were found dead of unknown cause on days 16 and 25 after randomization. Four animals (2 control, 2 treated) did not have evidence of MI on histological study and were excluded, which left 30 animals for analysis (16 control, 14 treated). The aortic valve was crossed in 27 animals, and cardiac output was successfully measured in 26 animals (13 control, 13 treated), with n=6 and n=7 in the small- and large-infarct groups, respectively. Angiography was technically satisfactory in 24 animals (12 control, 12 treated).
Effects of IGF-1/GH in the Conscious State
The effects of combined IGF-1 and GH (IGF-1/GH) treatment on systolic blood pressure and heart rate in the conscious state are shown in Fig 1⇓. At the time of randomization and 2 and 4 weeks later, there was no significant difference in systolic blood pressures between the two groups. At randomization, there was no difference in the resting heart rates between control and treated groups (384 and 384 bpm), although the resting heart rate in the treated group was mildly but significantly increased both at 2 weeks (390 versus 371 bpm) and 4 weeks (389 versus 361 bpm) (Fig 1⇓).
Effects of IGF-1/GH on Cardiac Output and Pressures Under Anesthesia
Under closed-chest, anesthetized conditions, the LV systolic pressures were not significantly different between treated and untreated groups. The LV end-diastolic pressures were elevated in both groups but not significantly different between groups (Table 1⇑). The heart rates were slower than under awake conditions. The mean aortic pressures were not different, and the small increase in LV dP/dtmax in the treated group, estimated from the fluid-filled catheter, was not statistically significant (Table 1⇑). None of the hemodynamic variables in Table 1⇑ was significantly related to infarct size.
The cardiac index (cardiac output/body weight) determined by the fluorescent microsphere technique in the six normal, closed-chest, anesthetized rats (without MI) averaged 215.8±35.4 mL·min−1·kg−1, which is within the range of values previously reported in normal anesthetized rats (208±26 mL·min−1·kg−1).27 In control (untreated) rats with MI, the average cardiac index was significantly reduced from normal by 29% (153.5 mL·min−1·kg−1; P<.001).
Body weight significantly increased by 22% in the treated group compared with controls (320.3 versus 362.7 g, P<.001). IGF-1/GH treatment significantly increased cardiac output (46%) compared with the controls, but because of the pronounced body growth in treated rats, the cardiac index was not significantly increased (Fig 2⇓). However, there was a significant interaction between treatment and infarct size, and all measures of cardiac output, including the cardiac index, were significantly increased in rats with large MIs by post hoc testing (Table 2⇓; Fig 2⇓); the cardiac index was 22% higher in these animals than in control rats with large MIs and approached the normal level (205 versus 216 mL·min−1·kg−1). There was a significant decrease of SVR (26%) in IGF-1/GH treated rats compared with controls (Table 1⇑).
Effects of IGF-1/GH on LV Volumes and Function by Angiography
Treatment caused a trend toward increased absolute LVEDV (P=.052); however, when normalized for the increased body weight induced by treatment, the ratios of LVEDV and LVESV to body weight were reduced in treated animals (Table 3⇓).
The average LV ejection fraction in all rats (treated and untreated) was markedly depressed compared with the ejection fraction of 70% previously reported in normal rats,4 but there was no significant difference in ejection fractions between treated rats (27.3%) and those receiving placebo (23%) (Table 3⇑). None of these variables showed an interaction between treatment and infarct size.
Effects of IGF-1/GH on Body, Organ, and Heart Weights
Treatment produced a significant increase in body weight and in selected organ weights (Table 4⇓). A small but significant increase in tibia length occurred (39.5 versus 40.3 mm). None of these changes was significantly correlated with infarct size.
Heart and ventricular weights, absolute and normalized to body weight and tibia length, are summarized in Table 5⇓. All of these variables (in both groups) were significantly correlated with infarct size. Treatment increased heart weight and the ratio of heart weight to tibia length compared with placebo in animals with large and small infarcts, but not when normalized for body weight. Treatment was associated with a trend toward increased LV weight (15%; P=.058) and a significant increase in the ratio of LV weight to tibia length but not of LV weight to body weight. RV weight and the ratio of RV weight to tibia length were significantly increased by treatment (the former by 56%), but the ratio of RV weight to body weight was not significantly increased.
Fig 3⇓ shows the relations between LV and RV weights and MI size (infarcts larger or smaller than the overall median of 39% of LV circumference). For LV weight, there was no significant interaction between treatment and infarct-size group (Fig 3⇓, left). However, for the right ventricle, there was a significant interaction between treatment and infarct-size group, and RV weight (Fig 3⇓, right) and the ratio of RV weight to tibia length but not to body weight were significantly increased in animals with large infarcts.
Other Postmortem Studies
MI size assessed by percentage of LV circumference demonstrated no significant differences between treated and control groups (40.9±9.4% versus 37.4±11.1%, respectively). Also, when percentage of LV cross-sectional area was used, there was no significant difference between treated and control groups (22.3±5.4% versus 20.5±4.2%, respectively). The range of infarct sizes also was similar between groups. Scar thickness was very similar (0.88±0.27 versus 0.88±0.28 mm, treated versus control).
LV Diastolic Pressure-Volume Relations
There was considerable scatter in the pressure-volume relations among individual hearts. When the average of fitted curves was used, the curves were nearly the same in the treated and placebo groups in animals with small infarcts (below the overall median) (Fig 4⇓). In animals with large infarcts, the LV diastolic pressure-volume relation tended to be shifted to the right compared with the placebo group (Fig 4⇓), but the overall difference was not statistically significant.
Analysis of the slopes of the semilog plots of the LV pressure-volume curves for the treated and control groups showed no significant difference in the passive compliance of the left ventricle between the two groups.
IGF-1 and GH Levels
Plasma IGF-1 levels averaged 193.1±62.1 ng/mL (range, 72.0 to 307.8 mg/mL) in control rats (n=16) and 434.8±154.5 ng/mL (range, 216.0 to 799.2 mg/mL) in treated rats (n=14) (P<.001). There were no detectable antibodies to hIGF-1. Because sampling took place more than 16 hours after the last injection, plasma GH levels were less than the standard in all but 1 treated animal, in which it was slightly elevated (372.5 pg/mL; range of standard, 2.0 to 250 pg/mL).
LV failure was present in this rat model of MI, as evidenced by a 29% reduction from normal values in the cardiac index, elevated LV end-diastolic pressure (17 mm Hg), and markedly reduced LV ejection fraction (23%) in the untreated animals. Circulatory congestion was not documented, however, which precludes a definition of congestive heart failure. In the present study, we wanted to determine whether or not the beneficial effects of short-term (2 weeks) IGF-1 treatment on LV remodeling and function that have been observed early after MI14 would also be present, together with the effects on cardiac output, at a later time with prolonged treatment. Therefore, IGF-1 and GH in combination were administered from the 4th to the 8th week after MI, when the infarct had healed and substantial LV remodeling had occurred. An important general finding was that the size of the MI significantly interacted with treatment effects on hypertrophy and cardiac output, which confirmed that statistical consideration of infarct size is necessary in considering late-treatment effects in this animal model.
Treatment Effects on Cardiac Output, Pressures, and Contractility
IGF-1 and GH are not cardiac selective, and treatment was found to cause substantial increases in body weight and kidney, liver, and spleen weights, with only a slight increase in tibia length.
It was observed that IGF-1/GH administration increased the cardiac output but not the cardiac index in the entire treated group. However, a significant interaction between the cardiac index and MI size was detected by two-way ANOVA, and post hoc testing revealed a substantial increase in the cardiac index that was limited to the treated group with large infarctions. Systolic and diastolic pressures were not different between treated and control groups, but the SVR was significantly lower in treated animals.
In normal rats with GH-secreting tumors, GH excess was associated with an increased cardiac index and decreased SVR.40 IGF-1 also has been reported to produce vasodilation when injected locally.41 42 This vascular effect of IGF-1 may involve an endothelium-dependent mechanism,43 possibly through increased synthesis of nitric oxide.44 The considerable decrease in SVR produced by IGF-1/GH treatment likely was a major factor in the observed increase of cardiac output in the treated group. Regardless of the mechanism, the increase in cardiac index in the treated animals with large infarcts to a near-normal level appears to be an important effect of IGF-1/GH.
Preliminary studies45 in normal rats described increased LV contractility assessed by high-fidelity micromanometry with IGF-1 or GH administration but an insignificant increase with the combination. Also, GH alone was reported to increase LV dP/dtmax and cardiac index at 6 weeks in conscious rats with MI,46 and IGF-1/GH treatment in the same conscious rat model but at a lower IGF-1 dose (2 mg·kg−1·d−1 versus 3 mg·kg−1·d−1) and a higher GH dose (2 mg·kg−1·d−1 versus 0.8 mg·kg−1·d−1) than we used also increased LV dP/dtmax and cardiac index.47 However, high-fidelity LV pressure measurements were not made in those studies, and the potential effects of infarct size within groups were not assessed. In the present study, increased myocardial contractility might have contributed to the increased cardiac output and augmented cardiac index in the treated large-infarct group, although the slight increase in LV ejection fraction with treatment was not significant. We did not measure high-fidelity LV pressure, however, and therefore the LV dP/dtmax values provide only an index of directional changes of this variable and not absolute values10 ; in addition, the small increase in LV dP/dtmax in the treated group was not statistically significant. Therefore, although other studies appear to indicate a positive inotropic effect of GH and IGF-1/GH treatment, whether or not myocardial contractility was substantially affected by IGF-1/GH treatment in the present study cannot be stated with certainty.
Treatment Effects on Cardiac Remodeling
With treatment, LV weight increased in these dysfunctional hearts in proportion to body weight, and the ratio of LV to body weight was not different in treated and control groups; thus, the increase may have been secondary either to direct growth-factor stimulation or to increased body weight alone. However, the ratio of LV weight to tibia length did increase significantly in the treatment group.
RV weight and the ratio of RV weight to tibia length showed no change in our previous study,14 but when IGF-1/GH was given later after MI, there was an increase in the ratio of RV weight to tibia length, and RV weight was augmented only in rats with large infarctions. This difference may reflect the longer duration and increased stimulus for hypertrophy from treatment acting on a chamber (the right ventricle) in which tissue loss from infarction does not occur, and might have contributed to RV compensation and improved cardiac index in the treated large-infarct group.
In normal rats, tibia length changes minimally beyond maturity; it remains independent of changes in body weight and correlates better than body weight with changes in cardiac cell size during aging.34 In catabolic states in rats, IGF-1 has been shown to increase body weight and to augment the fractional weights (grams per kilogram of body weight) of the stomach and gut, kidneys, spleen, and thymus48 ; however, fractional weights of heart, skin, liver, and lungs of the eviscerated carcass (primarily muscle and bone) were not altered by IGF-1 administration. Moreover, carcass fat content was not increased by IGF-1, which indicates that IGF-1 stimulated growth in lean tissue.48 Because of the disproportionate growth of some organs and the possibility of edema formation in the present study, heart weights were normalized by tibia length as well as body weight, since tibia length has been proposed as a better measure of the amount of lean body tissue than the total body weight.34
IGF-1/GH treatment produced only minor additional LV remodeling compared with controls in this late phase after MI, when infarct healing and initial LV remodeling were already present. The additional cardiac remodeling that did occur after IGF-1/GH treatment compared with controls resulted in significantly smaller ratios of LVEDV and LVESV to body weight in treated compared with control rats; whether or not this reduced relative LV-chamber dilation with treatment was favorable in this late phase after MI remains to be determined. This difference from our previous study14 may reflect a greater increase in body weight and additional time for remodeling to occur before treatment, not only in the treated but also in the untreated group.
IGF-1 and GH Levels
IGF-1 levels were elevated in treated rats, as expected. The GH dose on the morning of the final study was omitted because the timing of studies varied throughout the day (a number of animals were studied each day). Since the plasma level of injected GH peaks at 4 to 6 hours and becomes negligible by 12 hours,49 the finding of very low exogenous GH levels is predictable, even though the large, twice-daily doses undoubtedly induced GH peaks throughout the course of the study. The same dose of IGF-1 used in the present study was previously shown to produce hypertrophy in normal rats.14 The plasma levels of IGF-1 change little over a normal 24-hour period owing to the presence of high-molecular-weight IGFBPs. The plasma IGF-1 levels in the untreated MI group tended to be at the lower end of normal ranges reported in previous studies38 and slightly lower than in our previous report.14 The explanation for this finding is uncertain, but we used female rats, and IGF-1 levels are lower in young females than in males and fall with advancing age.50 Also, lower levels in the present study than in the prior investigation14 may relate to the more prolonged heart failure state in the present study; in this regard, IGF-1 levels have been suggested to be reduced in human subjects with chronic heart failure and to increase with treatment.51 The IGF-1 levels with combination therapy were only mildly increased above those we previously observed with IGF-1 treatment alone,14 and we found no evidence of induced antibodies to IGF-1. Thus, we did not observe in rats the striking augmentation of IGF-1 levels produced by concomitant IGF-1 and GH administration that is reported in humans.27
The ejection fraction is preload and afterload dependent, and although systemic arterial pressure and LV end-diastolic pressure measured at the time of angiography were not different between groups, use of a relatively load-independent marker of cardiac function such as the end-systolic pressure-volume relation would be desirable to assess myocardial contractility. Also, the fluid-filled catheter system used in these experiments, which was placed primarily for microsphere injections to determine cardiac output, does not have sufficient fidelity to measure absolute LV dP/dtmax accurately, and use of a high-fidelity catheter to measure dP/dt would provide better information on myocardial contractility. In the assessment of cardiac hypertrophy, measurement of cardiac dry weight is desirable, but this measurement was not feasible because of the need for cardiac fixation for histological assessment of infarct size. However, it seems unlikely that myocardial edema accounted for an important component of increased chamber weights associated with treatment at this time (late after infarction).
The present study provides evidence that exogenous administration of IGF-1 plus GH for 4 weeks beginning 1 month after MI causes only minor remodeling of the failing left ventricle compared with controls but substantially increases the cardiac index in rats with a large MI. The latter effect may have been largely due to an IGF-1/GH–mediated decrease in SVR in the treated group, although whether or not a positive inotropic effect was produced by treatment is uncertain because of methodological limitations. Given these effects, which appear to be independent of any remodeling effect, additional experimental studies on the actions of IGF-1 and GH on myocardial contractility and circulatory function in models of heart failure would seem desirable. It will also be important to determine whether or not a detrimental effect occurs when growth-factor treatment is discontinued.
IGF-1 and GH are nonspecific growth factors, and although it is anticipated that growth factors with more cardiac selectivity may become available, it is possible that in some other species, including humans, less marked body growth would occur. However, in patients with severe heart failure, weight loss and muscle atrophy are often significant problems, and the anabolic effect and trophic actions of nonspecific growth factors such as IGF-1 and GH on skeletal muscle might be beneficial in such patients. Thus, assessment of the role of growth-factor therapy in clinical heart failure may be warranted in the future.
Selected Abbreviations and Acronyms
|cardiac index||=||cardiac output/body weight|
|hGH||=||human growth hormone|
|IGF||=||insulin-like growth factor|
|IGFBP||=||insulin-like growth factor binding protein|
|LVEDV||=||left ventricular end-diastolic volume|
|LVESV||=||left ventricular end-systolic volume|
|MAP||=||mean arterial pressure|
|rhIGF-1||=||recombinant human insulin-like growth factor-1|
|SVR||=||systemic vascular resistance|
This study was supported by SCOR grant HL-53773 awarded by the National Heart, Lung, and Blood Institute. The authors acknowledge the statistical assistance of Elizabeth Gilpin. The authors also thank Farid Abdel-Wahhab for technical assistance and Cheryl Bugsch for manuscript preparation.
- Received November 2, 1995.
- Revision received December 27, 1995.
- Accepted January 2, 1996.
- Copyright © 1996 by American Heart Association
Anversa P, Loud AV, Levicky V, Guideri G. Left ventricular failure induced by myocardial infarction, I: myocyte hypertrophy. Am J Physiol. 1985;248:H876-H882.
Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial infarction: experimental observations and clinical implications. Circulation. 1990;81:1161-1172.
Ono S, Bhargava V, Ono S, Mao L, Hagan G, Rockman H, Ross J Jr. In vivo assessment of left ventricular remodelling after myocardial infarction by digital videocontrast angiography in the rat. Cardiovasc Res. 1994;28:349-357.
McKay RG, Pfeffer MA, Pasternak RC, Markis JE, Cone PC, Shoichiro N, Alderman JD, Ferguson JJ, Safian RD, Grossman W. Left ventricular remodeling after myocardial infarction: a corollary to infarct expansion. Circulation. 1986;74:693-702.
Pfeffer JM, Pfeffer MA, Fletcher PJ, Braunwald E. Progressive ventricular remodeling in rat with myocardial infarction. Am J Physiol. 1991;260:H1406-H1414.
Anversa P, Beghi C, Kikkawa Y, Olivetti G. Myocardial infarction in rats: infarct size, myocyte hypertrophy, and capillary density. Circ Res. 1986;58:26-37.
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.
Pfeffer MA, Braunwald E, Moyé LA, Basta L, Brown EJ Jr, Cuddy TE, Davis BR, Geltman EM, Goldman S, Flaker GC, Klein M, Lamas GA, Packer M, Rouleau J, Rouleau JL, Rutherford J, Wertheimer JH, Hawkins CM, on behalf of the SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction: results of the Survival and Ventricular Enlargement trial. N Engl J Med. 1992;327:669-677.
Schieffer B, Wirger A, Meybrunn M, Seitz S, Holtz J, Riede UN, Drexler H. Comparative effects of chronic angiotensin-converting enzyme inhibition and angiotensin II type 1 receptor blockade on cardiac remodeling after myocardial infarction in the rat. Circulation. 1994;89:2273-2282.
Duerr R, Huang S, Miraliakbar HR, Clark RG, Chien K, Ross J Jr. Insulin-like growth factor I enhances ventricular hypertrophy and function during the onset of experimental cardiac failure. J Clin Invest. 1995;95:619-627.
Powell-Braxton L, Hollingshead P, Warburton C, Dowd M, Pitts-Meek S, Dalton D, Gillett N, Stewart T. IGF-1 is required for normal embryonic growth in mice. Genes Dev. 1993;12B:2609-2617.
Engelmann G, Boehm KD, Haskell JF, Khairallah PA, Ilan J. Insulin-like growth factors and neonatal cardiomyocyte development: ventricular gene expression and membrane receptor variations in normotensive and hypertensive rats. Mol Cell Endocrinol. 1989;63:648-657.
Ito H, Hiroe M, Hirata Y, Motoyoshi T, Adachi S, Masayoshi S, Koike A, Nogami A, Marumo F. Insulin like growth factor-1 induces hypertrophy with enhanced expression of muscle specific genes in cultured rat cardiomyocytes. Circulation. 1993;87:1715-1721.
Fuller J, Mynett JR, Sugden PH. Stimulation of cardiac protein synthesis by insulin-like growth factors. Biochem J. 1992;282:85-90.
Wåhlander H, Isgaard J, Jennische E, Friberg E. Left ventricular insulin-like growth factor I increases in early renal hypertension. Hypertension. 1992;19:25-32.
Donohue TJ, Dworkin LD, Lango MN, Fliegner K, Lango RP, Benstein JA, Slater WR, Catanese VM. Induction of myocardial insulin-like growth factor-I gene expression in left ventricular hypertrophy. Circulation. 1994;89:799-809.
Isgaard J, Nilsson A, Vikman K, Isaksson OGP. Growth hormone regulates the level of insulin-like growth factor-I mRNA in rat skeletal muscle. J Endocrinol. 1989;120:107-112.
Gosteli-Peter MA, Winterhalter KH, Schmid C, Froesch ER, Zapf J. Expression and regulation of insulin-like growth factor-I (IGF-I) and IGF-binding protein messenger ribonucleic acid levels in tissues of hypophysectomized rats infused with IGF-I and growth hormone. Endocrinology. 1994;135:2558-2567.
Guler HP, Zapf J, Scheiwiller E, Froesch ER. Recombinant human insulin-like growth factor I stimulates growth and has distinct effects on organ size in hypophysectomized rats. Proc Natl Acad Sci U S A. 1988;85:4889-4893.
Zapf J, Schmid C, Guler HP, Waldvoger M, Hauri C, Futo E, Hossenlopp P, Binoux M, Froesch ER. Regulation of binding proteins for insulin like growth factor I (IGF) in humans: increased expression of IGF binding protein 2 during IGF-1 treatment of healthy adults and in patients with extrapancreatic tumor hypoglycemia. J Clin Invest. 1990;86:952-961.
Kupfer SR, Underwood LE, Baxter RC, Clemmons DR. Enhancement of the anabolic effects of growth hormone and insulin-like growth factor I by use of both agents simultaneously. J Clin Invest. 1993;91:391-396.
Fletcher PJ, Pfeffer JM, Pfeffer MA, Braunwald E. Left ventricular diastolic pressure-volume relations in rats with healed myocardial infarction: effects on systolic function. Circ Res. 1981;49:618-626.
Wicker C, Tarazi RC. Importance of injection site for coronary blood flow determinations by microspheres in rats. Am J Physiol. 1980;242:H94-H97.
Glenny RW, Bernard S, Brinkley M. Validation of fluorescent-labeled microspheres for measurement of regional organ perfusion. J Appl Physiol. 1993;74:2585-2597.
Omens JH, Rockman HA, Covell JW. Passive ventricular mechanics in tight-skin mice. Am J Physiol. 1994;35:H1169-H1176.
Yin RC, Spurgeon HA, Rakusan K, Weisfeldt ML, Lakatta EG. Use of tibial length to quantify cardiac hypertrophy: application in the aging rat. Am J Physiol. 1982;243:H941-H947.
Oh B-H, Ono S, Rockman HA, Ross J. Myocardial hypertrophy in the ischemic zone induced by exercise in rats after coronary reperfusion. Circulation. 1993;87:598-607.
Lieberman SA, Bukar J, Chen SA, Celniker AC, Compton PG, Cook J, Albu J, Perlman AJ, Hoffman AR. Effects of recombinant human insulin-like growth factor-I (rhIGF-I) on total and free IGF-I concentrations, IGF-binding proteins, and glycemic response in humans. J Clin Endocrinol Metab. 1992;75:30-36.
Oster MH, Fielder PJ, Levin N, Cronin MJ. Adaptation of the growth hormone and insulin-like growth factor-1 axis to chronic and severe calorie or protein malnutrition. J Clin Invest. 1995;95:2258-2265.
Fryburg DA. Insulin-like growth factor 1 exerts growth hormone- and insulin-like actions on human muscle protein metabolism. Am J Physiol. 1994;267:E331-E336.
Wu H-Y, Jeng YY, Yue C-J, Chyu K-Y, Hsueh WA, Chan TM. Endothelial-dependent vascular effects of insulin and insulin-like growth factor I in the perfused rat mesenteric artery and aortic ring. Diabetes. 1994;43:1027-1032.
Kiowski W, Stoschitsky K, Kim JH, Hussain MA, Froesch RE. Insulin-like growth factor-I causes endothelium-dependent arterial vasodilation. Circulation. 1994;90:1298. Abstract.
Cittadini A, Katz SE, Strömer H, Morgan JP, Douglas PS. Differential cardiac effects of growth hormone and IGF-1 in a new model of hormone excess in the rat: serial in vivo evaluation. Circulation. 1994;90(suppl I):I-17. Abstract.
Yang R, Bunting S, Gillett N, Clark R, Jin H. Growth hormone improves cardiac performance in experimental heart failure. Circulation. 1995;92:262-267.
Ballard FJ, Tomas FM, Read LC, Knowles SE, Owens PC, Lemmey AB, Martin AA, Wells JRE, Wallace JC, Francis GL. Effects of IGF-1 and IGF analogs on growth during catabolic states in rats. In: Spenser EM, ed. Modern Concepts of Insulin-Like Growth Factors. New York, NY: Elsevier Science Publishing Co Inc; 1991:617-627.