Author + information
- Received April 30, 2008
- Revision received July 31, 2008
- Accepted July 31, 2008
- Published online December 9, 2008.
- Susanne W.M. van den Borne, MD⁎,†,
- Satoshi Isobe, MD, PhD⁎,
- Johan W. Verjans, MD⁎,
- Artiom Petrov, PhD⁎,
- Dagfinn Lovhaug, MSc‡,
- Peng Li, MD, PhD⁎,
- H. Reinier Zandbergen, MD⁎,
- Youping Ni, MD, PhD⁎,
- Peter Frederik, PhD†,
- Jun Zhou, MD⁎,
- Bente Arbo, PhD‡,
- Astri Rogstad, PhD‡,
- Alan Cuthbertson, PhD‡,
- Salah Chettibi, PhD‡,
- Chris Reutelingsperger, PhD†,
- W. Matthijs Blankesteijn, PhD†,
- Jos F.M. Smits, PhD†,
- Mat J.A.P. Daemen, MD, PhD†,
- Faiez Zannad, MD, PhD, FACC§,
- Mani A. Vannan, MD, FACC⁎,
- Navneet Narula, MD⁎,
- Bertram Pitt, MD, FACC∥,
- Leonard Hofstra, MD, PhD† and
- Jagat Narula, MD, PhD, FACC⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Jagat Narula, Division of Cardiology, University of California, Irvine School of Medicine, 101 The City Drive, Building 53, Mail Route 81, Orange, California 92868-4080
Objectives The purpose of this study was to evaluate interstitial alterations in myocardial remodeling using a radiolabeled Cy5.5-RGD imaging peptide (CRIP) that targets myofibroblasts.
Background Collagen deposition and interstitial fibrosis contribute to cardiac remodeling and heart failure after myocardial infarction (MI). Evaluation of myofibroblastic proliferation should provide indirect evidence of the extent of fibrosis.
Methods Of 46 Swiss-Webster mice, MI was induced in 41 by coronary artery occlusion, and 5 were unmanipulated. Of the 41 mice, 6, 6, and 5 received intravenous technitium-99m labeled CRIP for micro–single-photon emission computed tomography imaging 2, 4, and 12 weeks after MI, respectively; 8 received captopril or captopril with losartan up to 4 weeks after MI. Scrambled CRIP was used 4 weeks after MI in 6 mice; the remaining 10 of 46 mice received unradiolabeled CRIP for histologic characterization.
Results Maximum CRIP uptake was observed in the infarct area; quantitative uptake (percent injected dose/g) was highest at 2 weeks (2.75 ± 0.46%), followed by 4 (2.26 ± 0.09%) and 12 (1.74 ± 0.24%) weeks compared with that in unmanipulated mice (0.59 ± 0.19%). Uptake was higher at 12 weeks in the remote areas. CRIP uptake was histologically traced to myofibroblasts. Captopril alone (1.78 ± 0.31%) and with losartan (1.13 ± 0.28%) significantly reduced tracer uptake; scrambled CRIP uptake in infarct area (0.74 ± 0.17%) was similar to CRIP uptake in normal myocardium.
Conclusions Radiolabeled CRIP allows for noninvasive visualization of interstitial alterations during cardiac remodeling, and is responsive to antiangiotensin treatment. If proven clinically feasible, such a strategy would help identify post-MI patients likely to develop heart failure.
- interstitial fibrosis
- radionuclide imaging
- heart failure
- coronary artery disease
Heart failure (HF) is evolving as 1 of the most important cardiovascular health problems worldwide. In the U.S. alone, approximately 5 million people suffer from manifest HF and more than 500,000 new cases are diagnosed every year (1). The syndrome of HF post-myocardial infarction (MI) is characterized by a relentless course of myocardial remodeling and functional deterioration (2–4), which continues to occur even after the initial causative injury has abated (5,6). In addition to the replacement fibrosis in the region of MI, interstitial fibrosis in the noninfarcted myocardium significantly contributes to the adverse remodeling and HF (7). In fact, presence of fibrosis remote from the infarct zone accounts for two-thirds of the fibrous tissue in the cardiomyopathic heart (7–10). The magnitude of myocardial fibrosis is correlated to the extent of ventricular dysfunction (11). It is conceivable that an ability to noninvasively detect the process of myocardial fibrosis would allow assessment of the likelihood of evolution of HF after MI.
Collagen production and fibrosis in the myocardium are associated with myofibroblastic proliferation (5). Myofibroblasts demonstrate up-regulation of angiotensin receptors and integrin moieties, which, in turn, may promote collagen genes and reduce metalloproteinase genes (12). Such effects of integrin up-regulation are prevented by abrogation of autocrine transforming growth factor-β signaling (12). The RGD peptide (containing the arginine-glycine-aspartate motif) that binds to integrins such as αvβ3 has been used to identify neovascularization in post-infarct animal models (13). Since integrin αvβ3 is associated with the supermature focal adhesions on the cell membrane of myofibroblasts, we hypothesized that appropriately labeled RGD probes should identify myofibroblasts in post-infarct myocardium (14). In addition, scrutiny of the pro-collagen I sequence revealed RGD binding domains, such as DDX, and could also constitute a target for RGD-based imaging. Therefore, uptake of RGD probe should indirectly represent the rate of fibrogenesis or collagen deposition.
In the present study, we used Cy5.5-RGD imaging peptide (CRIP) labeled with technetium (Tc)-99m, for feasibility of imaging the process of active myocardial fibrosis in a murine model of post-MI ventricular dysfunction. The noninvasive imaging ability of the radiolabeled probe was compared with echocardiographic parameters of left ventricular (LV) geometric changes and pathologic characterization of interstitial alterations. The fluorescent moiety of the targeting peptide allowed better histologic characterization of the probe localization. An antibody against the fluorescent moiety of CRIP was used to immunoelectron microscopically trace the localization of CRIP. In addition, in vitro experiments were performed for characterization of the CRIP binding to mature and pro-collagen.
Experimental myocardial infarction in mice
The experimental protocol was approved by the Institution Animal Care and Use Committee of the University of California, Irvine, School of Medicine. In 41 adult Swiss Webster male mice (age: 4 months; body weight: ∼50 g), MI was induced under pentobarbital (75 mg/kg) and isoflurane gas anesthesia (2.0% to 3.0%) using a stereomicroscope (Leica MZ FL III, Leica, Switzerland). For this purpose, animals were placed on a heating pad in the supine position, endotracheal intubation was performed under direct laryngoscopy, and mechanical ventilation was maintained with a small animal respirator (Harvard Apparatus, Holliston, Massachusetts) (tidal volume = 1.0 ml, rate = 120 breaths/min). After thoracotomy, the lateral branch of the left coronary artery was ligated with a 6.0-silk suture 3 to 4 mm below the tip of the left atrium. Successful ligation was verified by visual inspection of the LV apex for myocardial blanching, indicating interruption in coronary flow. The chest cavity was closed in layers with 6.0-silk, and the skin closed with 4.0-silk sutures. Animals were gradually weaned from the respirator.
For evaluation of serial changes in collagen synthesis by radiolabeled CRIP imaging, animals were divided into groups at 2 weeks (n = 6), 4 weeks (n = 6), and 12 weeks (n = 5) after infarction (Table 1). Two groups of 4 animals each were treated with either captopril (60 mg/kg/day) alone or in combination with losartan (captopril 30 mg/kg/day, losartan 10 mg/kg/day) dissolved in the drinking water, to evaluate if molecular imaging with CRIP would allow determination of efficacy of therapeutic intervention. Five unmanipulated control mice were subjected to CRIP imaging for comparison with the infarcted mice after a 4-week wait. In addition, 6 animals, 4 weeks after MI, were imaged with Tc-labeled scrambled peptide (scrambled CRIP) to ensure the specificity of CRIP. In the remaining 10 mice, nonradiolabeled CRIP was used for pathologic characterization of the target using fluorescence microscopy and immunoelectron microscopy.
All animals were subjected to extensive echocardiographic studies for the assessment of myocardial infarct size, LV cavity dimensions, and ventricular function.
Echocardiographic examination was performed under 2% isoflurane anesthesia for the assessment of infarct size, LV cavitary dimensions, and ventricular function before nuclear imaging. Echocardiograms were recorded with a commercially available ultrasound system (Sequoia, Siemens, Mountain View, California) using a 14-MHz linear probe (15L8, Siemens). An advanced high frame rate imaging technique, acoustic capture (Paragon, Siemens), was adopted for image acquisition allowing temporal resolution of 8 to 10 ms (frame rate, 100 to 120/s). B-mode images of LV parasternal long- and short-axis views at base, mid, and apical LV levels were digitally stored as movie loops in 2 to 3 cardiac cycle lengths. Still images of LV M-mode as well as pulsed-wave Doppler spectrum of mitral valve inflow and aortic valve forward flow were acquired at a sweep of 100 mm/s. Infarct and LV area were measured by tracing endocardial borders on long-axis images, and infarct-to-LV ratio was calculated. LV wall thicknesses and cavity dimensions were measured on LV M-mode spectrum as recommended by the American Society of Echocardiography guidelines (15), followed by calculation of percent fractional shortening and ejection fraction.
CRIP and scrambled CRIP: structure and radiolabeling
CRIP and scrambled CRIP were the kind gift of GE Healthcare, AS, Oslo, Norway. CRIP (AH110863) (Fig.1) is a 2.5-kD peptide conjugated to both a fluorescent cyanine dye Cy5.5 and the chelating agent cPN216; the latter is linked for radiolabeling of the peptide with 99mTc. The peptide comprises 10 amino acids, contains an RGD-motif, and has a bicyclic structure formed by a disulfide and a thioether bridge. The dye moiety and the chelating agent are conjugated to the peptide via the side-chain of lysine residue at the N- and C-terminal ends, respectively. A scrambled version of CRIP (AH112298) was also prepared, in which the order of the amino acids in CRIP was altered so that the sequence did not express an RGD-motif. The dye and the chelating agent were conjugated at the same positions.
For radiolabeling, 50 μg of tracer was dissolved in 50 μl of MeOH and then added to a freeze-dried kit. One milliliter of 99mTcO4− was added to the compound and left at room temperature for 20 to 30 min. Radiolabeling was confirmed by instant thin-layer chromatography with radiopurity of more than 90%. For nuclear imaging, 170 ± 18 MBq (4.6 ± 0.4 mCi) of Tc-CRIP was injected through the tail vein and imaging was performed after 3.5 h.
Micro–single-photon emission computed tomography (SPECT) imaging with Tc-CRIP and micro-computed tomography (CT)
Radionuclide imaging was performed using a dual-head micro-SPECT gamma camera with micro-CT (X-SPECT, Gamma Medica, Inc., Northridge, California) under isoflurane anesthesia. In vivo micro-SPECT images of the heart were acquired in a 64 × 64 scaffold, 32 steps at 120 s/step on a 140-keV photopeak of 99mTc with a 15% window using a low-energy, high-resolution, pinhole collimator. After SPECT acquisition, micro-CT images were acquired without moving the mice. The micro-CT used an X-ray tube operating at 50 kVp and 0.6 mA, and images were acquired for 0.5 s/view for 256 views in 360° rotation. The micro-SPECT images were converted to 256 × 256 scaffold and micro-CT studies were fused, allowing simultaneous scintigraphic and anatomic information in all tomographic scans in 3 different axes. After in vivo imaging, animals were sacrificed with an overdose of pentobarbital (150 mg/kg). Hearts were carefully explanted, and planar images of ex vivo heart were acquired for 15 min in a 128 × 128 matrix using a low-energy, high-resolution, pinhole collimator. Thereafter, hearts were cut into 3 bread-loaf slices (infarct, peri-infarct, and remote regions), using a mouse heart matrix (Zivic Laboratories, Inc., Pittsburgh, Pennsylvania). The quantitative radiotracer uptake was determined by a gamma scintillation counter (1480 Wizard 3, Wallac Co., Waltham, Massachusetts). Biodistribution studies for the lung, liver, spleen, and kidney uptake were also undertaken.
Histopathologic characterization of myocardial specimens
After the nuclear imaging studies, the apical (predominantly containing infarct tissue), mid myocardium (peri-infarct), and basal (predominantly containing noninfarcted remote tissue) myocardial slices were washed in phosphate-buffered saline (PBS), fixed overnight with 4% paraformaldehyde in PBS (pH 7.4 at 4°C), and stored in PBS with 0.02% sodium azide at 4°C until used. The specimens were further processed by dehydration in a graded series of ethanol for paraffin-embedding. The blocks were cut in 5-μm sections, transferred to Vectabond (Vector Laboratories, Burlingame, California) reagent-treated slides (Vector SP-1800, Vector Laboratories), dried overnight, and stored until ready for use.
Masson's trichrome staining was used to determine infarct size. After deparaffinization and rehydration, sections were placed in working Weigert's iron hematoxylin for 10 min and tepid water rinsed for 10 min. Tissue sections were then incubated in Biebrich scarlet-acid fuchsin solution for 5 min, differenced in phosphotungstic-phosphomolybdic acid for 5 min, and stained in aniline blue for 8 min. After rinsing with 1% gallic acetic acid, sections were dehydrated and mounted in permanent mount medium.
Sirius red staining eliminates cytoplasmic staining, revealing thin septa and collagen fibers clearly, and enabling quantitative morphometric measurements. Sections were treated in 0.2% aqueous phosphomolybdic acid for 5 min and subsequently incubated for 60 min with 0.1% sirius red F3BA (C.I. 35780, Polysciences, Northhampton, United Kingdom) in saturated picric acid and washed for 2 min with 0.01 N HCl. Sections were rinsed in 70% alcohol for 30 s, then dehydrated and mounted with a cover-slip, and the amount of collagen was quantified (16). The quality of collagen fibers was further investigated by sirius red polarization microscopy, allowing semiquantification of the mature, thick, tightly packed orange/red fibers and the newly formed, thin, loosely assembled fibers as yellow/green (17–20).
Alpha smooth muscle actin (ASMA) was used to determine myofibroblasts in the infarct area. After deparaffinization, rehydration, and blocking of endogenous peroxidase, sections were incubated for 2 h at room temperature using anti-ASMA (ASMA monoclonal antibody, Sigma, St. Louis, Missouri, dilution 1:2,000). After washing in PBS, sections were incubated with secondary antibody (rabbit antimouse horseradish peroxidase, 1:500, DAKO, Carpinteria, California) for 1 h at room temperature. Sections were briefly counterstained with hematoxylin.
For localization of the target sites for CRIP, nonradiolabeled probe was injected intravenously in animals with 2-week-old MI. After 3.5 h, mice were sacrificed and hearts were excised. Frozen sections of 5 μm were cut and rehydrated in PBS. Sections were incubated for 2 h at room temperature using anti-ASMA antibody (1:2,000, Sigma). After washing in PBS, sections were incubated with secondary antibody for 1 h at room temperature (donkey antimouse IgG FITC, Jackson ImmunoResearch Europe, Newmarket, Suffolk, United Kingdom). Sections were examined using a Nikon Eclipse E-800 microscope (Nikon Instruments Europe B. V., Amstelveen, the Netherlands).
In vitro studies for CRIP binding
Modified enzyme-linked immunoadsorbent assays were developed for defining the interaction of CRIP and collagen I and III (Sigma Aldrich, St. Louis, Missouri). In addition, a pro-collagen I peptide was custom constructed containing a DDX sequence (H2N-GPP-GKN-GDD-GEA-GKP-GR-COOH). This sequence is common for mouse and human pro-collagen I. The analysis was conducted using a Biacore 3000 system (Biacore, GE Healthcare, Uppsala, Sweden) that enabled detection of noncovalent interactions between CRIP and pro-collagen I peptide in real time, applying the physical principle of surface plasmon resonance. The analysis involved immobilization of synthetic peptide sequence on sensor chips CM5, followed by assessment of soluble CRIP or scrambled CRIP to the immobilized ligand. The change in refractive index at the chip surface layer, as the soluble components associated with the ligand and subsequently dissociated, was expressed as resonance units. To compensate for nonspecific background binding, an unmodified control surface was used. The change in resonance units at this surface was subtracted from the change in resonance units at the ligand surface.
Immunoelectron microscopy was performed according to previously described methods (21). In short, hearts were perfused with 0.2% glutaraldehyde in 2% paraformaldehyde and 0.1 mol/l phosphate buffer, pH 7.4. Hearts were excised and stored for 60 min in the previously mentioned solution before storage in 1% paraformaldehyde and 0.1 mol/l phosphate buffer, pH 7.4 at 4°C for at least 24 h. Tissue samples were stored in 2.3 mol/l sucrose in 0.1 mol/l phosphate buffer, and sections were cut and placed on a grid in methyl cellulose/2.3 mol/l sucrose solution (1:1). Anticyanine antibody was used (Acris Antibodies, Hiddenhausen, Germany) to ultrastructurally trace CRIP localization. After protein A gold labeling, sections were examined using a Philips CM100 microscope (Eindhoven, the Netherlands).
To determine the statistical significance of differences in quantitative scintigraphic data, echocardiographic parameters, and histopathologic observations, 1-way analysis of variance was performed followed by post-hoc Bonferroni test for multiple comparisons. Correlations were determined using linear regression analysis. Due to the small number of mice in each group, the assumptions of the 1-way analysis of variance and linear regression models should be regarded as exploratory rather than definitive. The p values of <0.05 were considered statistically significant.
CRIP: target affinity and localization characteristics
Probe localization characteristics were determined by fluorescence microscopy of myocardial tissue sections of hearts explanted 2 weeks after MI from animals injected with (nonradiolabeled) CRIP (Fig. 2). Uptake was observed predominantly in the myocardial infarct zone. Colocalization of ASMA in the same cells as CRIP-positive cells by fluorescence microscopy suggested the probe specificity for myofibroblasts. Similarly, immunoelectron microscopic examination was performed wherein intravenously administered CRIP in 2-week-old infarcts was localized by immunogold-labeled anticyanine antibody. Uptake was confirmed in the myofibroblasts, which were ultrastructurally identified by the presence of large amounts of rough endoplasmic reticulum. A small amount of internalized CRIP was seen within the myofibroblasts. No uptake of the probe was observed in the externalized collagen fibers. In vitro binding assays revealed CRIP uptake by activated αvβ3/5 with an affinity of 1-3nM, but not by typical collagen receptors such as α1β1 or α5β1 or platelet receptor αIIβ3. Also, CRIP did not bind to mature collagen type I or III fibers. However, a specific noncovalent interaction of CRIP was observed with a custom-made DDX-sequence containing pro-collagen I peptide (Fig. 2).
Radionuclide imaging with Tc-CRIP
The infarct size in all untreated animals was similar (Table 2). The LV cavity dimensions (both end-systolic and diastolic) were greater at 2, 4, and 12 weeks post-MI, and the percent fractional shortening and ejection fraction decreased post-MI compared with those in normal control animals. In vivo micro-SPECT imaging demonstrated Tc-CRIP uptake in the infarct area (Fig. 3). The use of micro-CT identified precise localization of the radioactivity in the cardiac region and allowed differentiation of the uptake in apical myocardium from radiotracer sequestration in the liver. Ex vivo planar images of the explanted heart specimens confirmed the results of in vivo imaging and demonstrated maximum uptake in infarct and border zones. Minimal increase in uptake was seen in the remote area (Fig. 4). In contrast to CRIP, scrambled CRIP did not show uptake in the infarcted myocardium, showing that CRIP uptake is RGD-dependent and -specific. Similarly, no Tc-CRIP uptake was seen in the unmanipulated hearts.
Uptake in the infarcted region was maximal in the 2-week-old MI. There was statistically significant reduction in uptake in 4- and 12-week-old infarcts (Fig. 4). Quantitative Tc-CRIP uptake in the myocardial tissue specimens was calculated and represented as percent injected dose/g (%ID/g) of tissue. It can be presumed that the quantitative uptake represents the extent of the process of myocardial fibrosis. As observed in the in vivo and ex vivo images, the maximum radiotracer uptake was observed in the infarcted myocardium (%ID/g: 2.75 ± 0.46%) at 2 weeks, which was 5-fold higher than the control myocardial specimens (0.59 ± 0.19%; p < 0.0001) (Fig. 4B). This uptake was significantly higher compared with specimens from 4 (2.26 ± 0.17%; p < 0.0038) and 12 (1.74 ± 0.24%; p < 0.0001) weeks after MI. The CRIP uptake in the remote myocardium was higher than that in the control myocardial specimens, and it increased significantly by 12 weeks (1.04 ± 0.13%), as compared with uptake in control myocardial specimens (0.68 ± 0.16%; p = 0.001). The uptake in the infarct area had no correlation with MI size determined either by akinetic echo segments or Masson trichrome staining (r = −0.17, −0.12 respectively; p = NS). The nontarget organ distribution of Tc-CRIP demonstrated the kidney to be the major organ of radiation burden (and excretion) (Online Appendix). The tracer clearance was fast, and only 8% of the tracer could be recovered in circulation at 2 h as compared with 5 min after administration; CRIP T1/2 was calculated as 28 min.
Two additional groups of 4 animals each were treated with either captopril alone or in combination with losartan. Upon ultrasound examination, the neurohumoral antagonists prevented ventricular dilation and preserved ventricular function (Table 2). The CRIP uptake at 4 weeks after MI was significantly reduced after captopril (1.78 ± 0.31%; p = 0.0004) or captopril + losartan (1.13 ± 0.28%; p = 0.0001) treatment compared with uptake in untreated animals (Fig. 5).
The histologic sections demonstrated large infarcts involving the anterior LV wall (Fig. 6); infarct size was similar in all animals. During the 3 months following infarction, the LV wall demonstrated significant thinning and the extent of inflammation gradually subsided. Myofibroblasts, expressing ASMA, were present at 2 weeks post-MI in the infarct area and decreased over time. Replacement and interstitial collagen was analyzed by picrosirius red polarization microscopy of myocardial specimens (Fig. 7) obtained from infarct and remote regions. Although the total collagen content in the infarct region remained similar over time, the thin, newly formed (yellow/green) collagen fibers reduced and correlated with the radiotracer uptake (r2 = 0.34; p = 0.05) (Figs. 7A, 7C, and 7E). In the remote region, overall collagen deposition was markedly lower compared with that in the infarct region (Figs. 7A and 7B). However, both collagen content and (yellow/green) thin collagen fibers increased in the remote region over time (Fig. 7D). There was a direct correlation between tracer uptake and the thin collagen fiber deposition (r2 = 0.45; p = 0.001) (Fig. 7F). The direct correlation of CRIP uptake with ASMA-positive myofibroblasts and thin collagen fibers suggests that collagen production is decreased in the infarct zone over time.
The present study demonstrates the feasibility of noninvasive assessment of interstitial alterations in post-infarction ventricular myocardium by Tc-99m labeled CRIP. The lack of binding of scrambled CRIP (which carries a deranged RGD-motif) confirmed that the CRIP binding is RGD-dependent and specific. Fluorescence and immunoelectron microscopy confirmed the binding of CRIP to the myofibroblasts. The CRIP uptake paralleled newly formed, thin, yellow/green collagen fiber production, suggesting that the CRIP uptake should indirectly represent the rate of collagen deposition. CRIP did not bind to the mature collagen fibers.
The uptake of CRIP in the infarct area was significantly higher at 2 weeks compared with that at 4 or 12 weeks after MI; the uptake almost completely resolved by 52 weeks in the infarct area (data not shown). The time-response curve of CRIP uptake in post-MI hearts may parallel the expression dynamics of αvβ3, as observed in hepatic stellate cells (22). Myofibroblasts evolve and are activated by inflammatory cytokines elicited in the area following the insult. Collagen and other matrix molecules are produced, and integrins are up-regulated to attach the cells to the matrix for survival and proliferation. The expression of β3 integrins is associated with angiogenesis in the peri-infarct zone and peaks around 7 days after MI as part of the remodeling process. Subsequently, as inflammation resolves, collagen fibrils are cross linked by transglutaminase activity, no more ligands are available for αvβ3 binding, and the cells disappear through apoptosis or anoxis (22). Therefore, after 12 months there are large amounts of collagen, but no more integrins for CRIP to bind to; in other words, there is fibrosis, but little fibrogenesis or myofibroblasts in mouse infarcts. However, since myofibroblasts have been reported up to 20 years after MI in well-healed human infarcts (23), the binding characteristics of CRIP may be different in humans than in rodent hearts.
Unlike the infarcted zone, CRIP uptake in the remote myocardium was higher at 12 weeks. It is well established that alterations in interstitial collagen can significantly influence the size and shape of the cardiac chamber as well as ventricular function and hence, ventricular remodeling (24). Excessive collagen deposition or pathologic fibrosis contributes to LV dysfunction and poor prognosis in MI patients by inducing myocardial stiffness, promoting development of arrhythmias, and adversely affecting systolic function (25).
The myocardium contains an endogenous renin-angiotensin system, and activation of the renin-angiotensin-aldosterone axis results in cardiac fibroblast proliferation and fibrosis (26,27). This response is transduced by angiotensin receptors, which are far more abundant on cardiac fibroblasts than on myocytes (28). Angiotensin II acts through the up-regulation of additional fibrogenic growth factors, which mediate or augment the effects of angiotensin II, including transforming growth factor β1. Up-regulation of angiotensin production, angiotensin II type 1 receptors, and increased collagen messenger ribonucleic acid in myofibroblasts is associated with healing infarct scars (29), and antiangiotensin therapy may prevent adverse myocardial remodeling. The uptake of Tc-CRIP was markedly reduced by captopril alone and more so in combination with losartan. Although angiotensin II levels decrease initially with angiotensin-converting enzyme inhibitor therapy (30,31), their levels may rise gradually over time due to alternative conversion pathways. Therefore, therapy with a combination of an angiotensin-converting enzyme inhibitor with an angiotensin receptor blocker acts synergistically and may have a benefit greater than that with either agent used alone. The present study demonstrates the relatively superior effect of combination therapy in reducing fibrogenesis compared with captopril therapy alone. Although the OPTIMAAL (Optimal Therapy in Myocardial Infarction with the Angiotensin II Antagonist Losartan) and the VALIANT (Valsartan in Acute Myocardial Infarction) studies did not substantiate the concept of combination therapy in post-MI patients (32,33), the Val-HeFT (Valsartan Heart Failure Trial) and CHARM (Candesartan in Heart Failure–Assessment of Reduction in Mortality and Morbidity) studies have demonstrated the superior efficacy of combination therapy in reduction of morbidity in HF patients with (CHARM Addition) (34) or without (Val-HeFT) (35) improvement in survival. Since both agents are known to favorably alter prognosis in patients with HF and the present study demonstrates a decrease in collagen deposition after captopril with or without losartan treatment, it is possible that the decrease in collagen deposition is a favorable phenomenon. Alternatively, we hypothesize that the impact of antiangiotensin therapy on the infarct zone may be relatively lower as compared with that in the remote region, where all collagen formation may get abrogated. As such, evaluation of the impact of therapy at 12 weeks could have offered more information.
Since LV remodeling post-MI is the leading cause of HF and is a determinant of morbidity and mortality (3,36), it is important to predict the likelihood of occurrence of cardiac remodeling. It is possible to detect the occurrence of collagen formation post-MI by measuring pro-collagen I and III in serum. The measurements of these markers have been shown to provide independent information pertaining to LV function (37,38) and survival (39,40) in patients with HF. However, these markers are not specific and may not be of value in patients with other conditions associated with collagen formation, such as arthritis (41). The potential value of radiolabeled CRIP is its ability to noninvasively localize at the site of interstitial alterations and collagen formation in the myocardium, and therefore may provide information about the evolution of HF. Our pilot clinical study with a radiolabeled RGD peptide in patients at 3 and 8 weeks after MI has shown the feasibility of interstitial imaging. In these patients, a 1-year follow-up magnetic resonance imaging study confirmed the extent of fibrosis detected earlier by the radiolabeled RGD uptake (42).
The authors thank Hans Duimel for his help with the immunoelectron microscopy.
For a table on the tracer uptake in organs and blood, please see the online version of this article.
Dagfinn Lovhaug and Drs. Arbo, Rogstad, Cuthbertson, and Chettibi, who prepared the tracer for the imaging studies, are employees of GE Healthcare. Dr. Pitt is a consultant to Pfizer, Merck, Takeda, AstraZeneca, Synvista, Novartis, and Nile Therapeutics, but has no conflicts directly with the project. Dr. van den Borne was supported by a grant from the Van Walree Fund of the Royal Netherlands Academy of Arts and Sciences. Dr. Verjans was partially supported by the DiPalma-Brodsky research grant to Dr. Jagat Narula. CRIP and scrambled CRIP were kindly provided to Dr. Jagat Narula by GE Healthcare, Oslo, Norway. Drs. van den Borne and Isobe contributed equally to this study. Joel S. Karliner, MD, served as Guest Editor for this article.
- Abbreviations and Acronyms
- alpha smooth muscle actin
- Cy5.5-RGD imaging peptide
- computed tomography
- heart failure
- left ventricle/ventricular
- myocardial infarction
- phosphate-buffered saline
- single-photon emission computed tomography
- Received April 30, 2008.
- Revision received July 31, 2008.
- Accepted July 31, 2008.
- American College of Cardiology Foundation
- Thom T.,
- Haase N.,
- Rosamond W.,
- et al.
- Pfeffer M.A.,
- Braunwald E.
- Pfeffer J.M.,
- Pfeffer M.A.,
- Braunwald E.
- Cleutjens J.P.,
- Blankesteijn W.M.,
- Daemen M.J.,
- Smits J.F.
- Beltrami C.A.,
- Finato N.,
- Rocco M.,
- et al.
- Asano Y.,
- Ihn H.,
- Yamane K.,
- Jinnin M.,
- Mimura Y.,
- Tamaki K.
- Schiller N.B.,
- Shah P.M.,
- Crawford M.,
- et al.
- MacKenna D.A.,
- Omens J.H.,
- McCulloch A.D.,
- Covell J.W.
- Zhou X.,
- Jamil A.,
- Nash A.,
- et al.
- Jugdutt B.I.,
- Amy R.W.
- Whittaker P.,
- Boughner D.R.,
- Kloner R.A.
- Weber K.T.
- Gray M.O.,
- Long C.S.,
- Kalinyak J.E.,
- Li H.T.,
- Karliner J.S.
- Sun Y.,
- Weber K.T.
- Schieffer B.,
- Wirger A.,
- Meybrunn M.,
- et al.
- Dickstein K.,
- Kjekshus J.
- White H.D.,
- Braunwald E.
- Poulsen S.H.,
- Host N.B.,
- Jensen S.E.,
- Egstrup K.
- Verjans J.W.,
- Wolters S.L.,
- Lax M.,
- et al.