Author + information
- Received March 18, 2005
- Accepted April 20, 2005
- Published online August 16, 2005.
- Wangde Dai, MD,
- Loren E. Wold, PhD,
- Joan S. Dow, BS and
- Robert A. Kloner, MD, PhD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Robert A. Kloner, The Heart Institute, Good Samaritan Hospital, University of Southern California, 1225 Wilshire Boulevard, Los Angeles, California 90017
Objectives We determined whether collagen implantation could thicken the infarcted left ventricular (LV) wall and improve LV function.
Background We hypothesized that thickening the infarcted wall by using collagen might result in some benefits that are similar to what previously had been reported when the infarcted wall was thickened with cells.
Methods Fischer rats with one-week-old myocardial infarcts were injected with collagen or saline (100 μl) into the scar (n = 12 each group). Six weeks later, LV angiograms, hemodynamics, and regional myocardial blood flow were assessed. The hearts were processed for measurements of postmortem LV volume and histology.
Results Collagen injection significantly increased scar thickness (719 ± 26 μm) compared with the saline-treated group (440 ± 34 μm, p = 2.6 × 10−6). By LV angiography, stroke volume was significantly larger in the collagen-treated group (163 ± 8 μl) than in the saline-treated group (129 ± 6 μl, p = 0.005), and LV ejection fraction was also greater in the collagen-treated group (48.4 ± 1.8%) than in the saline-treated group (40.7 ± 1.0%, p = 0.002). Analysis of regional wall motion demonstrated paradoxical systolic bulging in 5 of 10 saline-treated rats that averaged 20.3 ± 2.6% of the LV diastolic circumference, but in none of the 11 collagen-treated rats (p = 0.012). The LV end-diastolic and end-systolic volumes were 319 ± 12 μl and 190 ± 7 μl in the saline-treated group, respectively. There was a trend for larger LV end-diastolic volumes (343 ± 23 μl), but smaller end-systolic volumes (180 ± 16 μl) in the collagen-treated group.
Conclusions This study shows that collagen injection thickens an infarct scar and improves LV stroke volume and ejection fraction, and prevents paradoxical systolic bulging after myocardial infarction.
After a proximal coronary artery occlusion, a wave front of necrosis progresses from the subendocardium to the subepicardium of the left ventricle (LV) within the area at risk (1). Approximately 15 min after severe ischemia, necrosis starts in the subendocardium and becomes transmural by about six h of occlusion. Within the first weeks of an extensive myocardial infarction (MI), the necrotic tissue stretches and thins as the collagen matrix holding the myocytes together disintegrates, allowing the necrotic myocytes to slip past each other (slippage) (2). This phenomenon is termed infarct expansion and results in some degree of early LV dilation. Over time, the infarct and noninfarct zones continue to thin and stretch; within the noninfarcted wall, eccentric hypertrophy develops over weeks to months as sarcomeres are added to the myocytes in a longitudinal fashion. This contributes further to LV dilation and, eventually, decompensation with congestive heart failure. These topographical changes that occur within the LV after acute MI—infarct expansion, concentric hypertrophy, and LV dilation—have been described as LV remodeling after acute MI.
Paradoxical systolic bulging caused by infarct expansion and thinning can impair LV systolic function (3). Recently, we have observed that injection of neonatal cardiomyocytes into the thinned infarct wall of rats subjected to acute MI increased scar thickness and improved regional wall motion—particularly reducing the degree of dyskinesis of the infarct zone (3). The benefit of preventing paradoxical systolic bulging that was observed might be due to the contractile potential of the implanted myocytes or related to passive thickening of the scar. By simply thickening and toughening the infarct scar, paradoxical systolic bulging should be reduced, wall stress should be reduced, and forward cardiac output improved. In the present study, we determined the feasibility and tolerability of injecting commercially available collagen into an infarct wall and whether collagen implantation can have many of the same benefits as cell transplantation, including: improving the thickness of the infarcted LV wall, improving global LV function, and preventing paradoxical systolic bulging.
All of the experiments were performed in accordance with the “Guide for the Care and Use of Laboratory Animals” (National Institutes of Health Publication No. 85-23, National Academy Press, Washington, DC, revised 1996) and approved by the Institutional Animal Care and Use Committee. The Heart Institute in Good Samaritan Hospital is accredited by the American Association for Accreditation of Laboratory Animal Care.
Collagen for implanting
Zyderm collagen implant (Zyderm type 2) was obtained from INAMED Corp. (Santa Barbara, California). It is a sterile compound composed of highly purified bovine dermal collagen (65 mg/ml) that consists of 95% collagen I and 5% collagen III. Zyderm collagen is used clinically for correction of contour deficiencies of soft tissue.
Model of MI and collagen injection
Myocardial infarction was performed by ligation of the proximal left coronary artery through a left thoracotomy under anesthesia with ketamine (75 mg/kg intraperitoneal) and xylazine (5 mg/kg intraperitoneal) in female Fischer CDF rats (Charles River Laboratories, Wilmington, Massachusetts) as described previously (3). One week later, the rats were re-anesthetized as described above. The hearts were exposed through a second thoracotomy, and collagen (∼100 μl Zyderm type 2 = 6.5 mg highly purified bovine dermal collagen; n = 12) or saline (∼100 μl, n = 12) was injected directly into the infarct area with a 28-gauge needle attached to an insulin syringe. The rats were allowed to recover and housed in a clean laboratory room with a temperature-controlled environment under a 12-h light-dark cycle and fed with free access to food and water for 6 weeks.
In vivo ventriculography
Left ventricular contrast angiography was performed with a XiScan 1000 C-arm X-ray system (XiTec Inc., East Windsor, Connecticut; 3-inch field of view). Rats were anesthetized at six weeks after injection, and a catheter was inserted into the left jugular vein. After the injection of 1 ml nonionic contrast into the left jugular vein, video images (anterior-posterior and lateral projections) were acquired on half-inch super-VHS videotape at 30 frames/s under constant fluoroscopy. The video images were analyzed off-line, blindly, to calculate LV volumes in systole and diastole. All parameters were averaged over three consecutive cycles in both projections. Ejection fraction (EF) (%) was calculated as: 100 × (volume in diastole − volume in systole)/volume in diastole, and averaged over both projections.
Assessment of paradoxical systolic bulging
Tracings obtained from the LV angiograms during end-diastole and end-systole of the same cardiac cycle of three consecutive beats were superimposed on transparent film and were compared between groups in both the anterior-posterior and lateral views. Bulging was evident if the tracing from end-systole was not confined within that of end-diastole. In order to calculate the size of paradoxical LV systolic bulging, the length of total LV diastolic circumference and circumferential length of the bulging segment were measured with computerized planimetry.
For hemodynamic measurements, a 2-F high-fidelity, catheter-tipped micromanometer (model SPR-869, Millar Inc., Houston, Texas) was advanced into the ascending aorta and LV through the right carotid artery to record aorta and LV hemodynamic parameters.
Regional myocardial blood flow (RMBF)
In order to measure RMBF, 103ruthenium-labeled radioactive microspheres (∼500,000) were injected directly into the LV, and a reference blood sample was withdrawn simultaneously from an arterial catheter (0.361 ml/min) for 1 min. Radioactivity in the scar tissue, the noninfarcted myocardium, and in the reference blood sample was counted in a multi-channel pulse-height analyzer (model ND62, Nuclear Data, Schaumburg, Illinois). After correction for background, RMBF was calculated as the ratio of counts in the tissue and the reference blood sample multiplied by pump flow (0.361 ml/min) and divided by the weight of the tissue.
Postmortem LV volumes, wall thickness, infarct sizes of hearts
Intravenous injection of 0.6 ml 50% Unisperse blue (Ciba Geigy, Hawthorne, New York), a suspension of blue particles, was performed to stain the capillaries in the collagen injection area. Under deep anesthesia, the rats were euthanized with 2 mEq potassium chloride intravenously to arrest the heart in diastole, and a cardiectomy was performed. The LVs were pressure-fixed (pressure = 13 cm water column) in formalin, and the volumes of LV were measured by filling the cavity with water and weighing, repeated three times.
Formalin-fixed hearts were cut into three transverse slices. The middle slice was embedded in paraffin for histology, and the other two slices were used for RMBF analysis. The paraffin-embedded tissue was sectioned (5-μm thickness) and stained with hematoxylin and eosin and picrosirius red. Histological images of the stained sections were traced with computerized planimetry, and the following parameters were measured: 1) scar thickness (average of five equidistant measurements), right ventricular free wall thickness, and septum thickness (average of three equidistant measurements); 2) epicardial circumference and endocardial circumference; and 3) circumference occupied by noninfarcted muscle and collagen. Septal fibrosis was assessed through calculating collagen deposition area in the septum. Picrosirius red staining slices, in which the collagen stained red, were used to quantify the collagen deposition area. We used a 10 × objective lens with a 100-square grid (10 × 10). Squares were counted either positive if they contained red color or negative if they did not. The degree of collagen deposition was expressed as a percentage of positive squares. Verhoeff's stain for elastic tissue staining was performed to detect the elastic fibers within the collagen implant.
All data are presented as mean ± SEM. Comparison between groups was made by Student ttest, or Fisher exact test, where appropriate. Results were considered statistically significant if p < 0.05.
A total of 24 rats initially were included in this study. One rat in the phosphate-buffered saline group died after the chest was closed and the intubation tube was removed during the second surgery. One rat in the collagen group died one day after collagen injection. Of the 22 successful rats, one rat in the saline group was excluded because no evidence of MI was found in the heart. The LV angiographic analysis, postmortem volume measurements, and histological analyses were performed in 11 collagen-treated rats and 10 saline-treated rats.
LV volume and EF by angiography
In the saline group, LV end-diastolic and end-systolic volumes were 319 ± 12 μl and 190 ± 7 μl, respectively. The stroke volume was 129 ± 6 μl. The EF was 40.7 ± 1.0%. In the collagen-treated group, there was a trend for larger LV end-diastolic volumes (343 ± 23 μl) and smaller end-systolic volumes (180 ± 16 μl) and significantly higher stroke volume (163 ± 8 μl, p = 0.005) and greater EF (48.4 ± 1.8%, p = 0.002) compared with the saline group (Fig. 1).
Paradoxical LV systolic bulging
In the saline-treated group, 5 of 10 rats showed paradoxical LV systolic bulging (dyskinesis). The extent of the paradoxical LV systolic bulging, expressed as percentage of total LV diastolic circumference, was 20.3 ± 2.6% in these saline-treated rats. Paradoxical LV systolic bulging was absent in all 11 rats treated with collagen (p = 0.012). Collagen injection significantly prevented asynchronous wall motion after MI.
No significant differences were noted in heart rate, systolic and diastolic blood pressure, LV positive change in pressure over time (dp/dt), and negative dp/dt between the two groups (Table 1).
Although RMBF in the scar tissue and noninfarct tissue was comparable between the two groups (Table 1), there was a nonsignificant trend for lower RMBF in scar tissue in the collagen-treated group (0.47 ± 0.10 ml/min/g, n = 9) compared with the saline group (0.69 ± 0.15 ml/min/g, n = 8, p = 0.22).
Postmortem LV volumes, wall thickness, noninfarct and infarct LV wall sizes, expansion index, and histology
Postmortem LV volume was comparable between the saline-treated group (337 ± 17 μl) and the collagen-treated group (325 ± 20 μl, p = 0.64). Collagen injection significantly increased scar thickness (719 ± 26 μm) compared with the saline group (440 ± 34 μm, p = 2.6 × 10−6) (Table 1), but did not increase collagen deposition in the septum and did not change right ventricular free wall thickness (Table 1).
The length of the epicardial and endocardial circumferences was comparable between the two groups (Table 1). The percentage of total LV circumference composed of collagen was significantly larger, whereas the noninfarcted LV wall, also expressed as a percentage of the total LV circumference, was significantly smaller in the collagen-treated group (Table 1). The expansion index as defined by Hochman and Choo (4), which is expressed as LV cavity area/total LV area × septum thickness/scar thickness, was significantly reduced in the collagen-treated group (Table 1).
Implanted collagen in the scar tissue was easily identified histologically as homogeneous bright red tissue on picrosirius red staining and no bundle formation of newly generated host collagen (Figs. 2 and 3).Hematoxylin and eosin staining demonstrated no cellular infiltration (inflammation) within the homogeneous collagen implant (Fig. 4).Verhoeff's stain for elastic tissue staining found that no elastic fibers were present within the homogeneous collagen implant (Fig. 5).The in vivo histologic appearance of the collagen implant was virtually identical to the in vitro appearance (Fig. 3D). Angiogenesis was not observed, because no blue particles appeared in the implanted collagen area (Figs. 3 and 4).
This study demonstrates for the first time that collagen implanted into the scar of a MI remained in the infarct zone for up to six weeks and did not induce an inflammatory response or immuno-rejection. Injection of collagen significantly increased scar thickness, prevented stretch of the noninfarcted zone, and improved LV stroke volume and EF after MI compared with saline injection. Improvement in stroke volume and EF compared with controls might have been largely due to a prevention of paradoxical systolic bulging observed in the collagen group.
A novel approach to preserve cardiac function after MI is implanting biomedical materials into an infarcted area as an internal support. Christman et al. (5) injected fibrin glue into one-week-old myocardial infarcts as an internal support and found this treatment preserved cardiac function at five weeks after injection; however, they did not know the exact mechanism by which fibrin glue preserved cardiac function. Fibrin glue might have worked by inducing neovasculature formation in the infarct area and/or attenuating the decrease in infarct wall thickness. In our present study, collagen injection attenuated the decrease in infarct wall thickness without neovasculature formation in the collagen implant. Thus, our data support the hypothesis that internal support of the infarct wall by biomedical materials can preserve cardiac function by thickening the infarct wall and without neovascularization. The advantages of collagen injection are that commercially prepared collagen is readily available, it can be injected early (before infarct expansion is complete), it is relatively inert (will not induce a rejection phenomenon), and is easy to inject into the scar. Compared with cell transplantation, collagen injection did not improve systolic and diastolic blood pressure and positive and negative dp/dt. The explanation might be that collagen injection did not affect the LV contractile ability, whereas the cell transplantation did. Although this study shows that collagen can thicken the scar, improve EF and stroke volume, and prevent paradoxical systolic bulging, the long-term effects need further investigation. Many questions remain unknown, such as how long the collagen can stay in the infarct area and whether the collagen is ultimately colonized by host connective tissue.
Another potential benefit of collagen injection is that by preventing scar shrinkage it might limit noninfarct LV stretching and eccentric hypertrophy of the noninfarcted LV. Roberts et al. (6) found that the infarcted tissue contracted, whereas noninfarcted tissue expanded during repair of the infarct 21 days after MI in rats. It is possible that scar contraction contributed to compensatory expansion of noninfarcted myocardium (eccentric hypertrophy) with eventual volume overload. This compensatory hypertrophy of noninfarcted myocardium after MI in the rat model was also observed by other groups (7). Holmes et al. (8) observed scar shrinkage after coronary ligation in the pig. The infarct scar contraction and the compensatory hypertrophy of noninfarcted myocardium might contribute to diastolic dysfunction caused by abnormal ventricular chamber stiffness (9). In the present study, the percent of the LV circumference composed of collagen was, of course, greater in the collagen group, whereas the noninfarcted myocardial circumference was significantly greater in the saline group than in the collagen injection group. Our data support the concept that collagen injection might have limited the development of eccentric (length-wise) hypertrophy of the noninfarcted myocardium.
In conclusion, our data suggest that collagen injection into the scar after MI can thicken the scar, improve LV stroke volume and EF, and limit paradoxical systolic bulging compared with saline-treated animals. Collagen injection has many of the same benefits as cell transplantation therapy, but might be a lower cost and technically easier option. This study suggests that implantation of biomedical materials into an MI as an internal support can provide a novel approach to preserve cardiac function after MI.
This study was supported in part by National Heart, Lung, and Blood Institute Grant R01-HL073709.
- Abbreviations and Acronyms
- change in pressure over time
- ejection fraction
- left ventricle/ventricular
- myocardial infarction
- regional myocardial blood flow
- Received March 18, 2005.
- Accepted April 20, 2005.
- American College of Cardiology Foundation
- Connelly C.,
- Vogel W.M.,
- Hernandez Y.M.,
- Apstein C.S.
- Whittaker P.,
- Boughner D.R.,
- Kloner R.A.
- Müller-Ehmsen J.,
- Peterson K.L.,
- Kedes L.,
- et al.
- Hochman J.S.,
- Choo H.
- Rubin S.A.,
- Fishbein M.C.,
- Swan H.J.C.
- Holmes J.W.,
- Yamashita H.,
- Waldman L.K.,
- Covell J.W.
- Sun Y.,
- Weber K.T.