Secondary Logo

Journal Logo

Tissue Engineering\Biomaterials

Scaffold-Based Transplantation of Vascular Endothelial Growth Factor—Overexpressing Stem Cells Leads to Neovascularization in Ischemic Myocardium but Did Not Show a Functional Regenerative Effect

von Wattenwyl, Robert; Blumenthal, Britta; Heilmann, Claudia; Golsong, Peter; Poppe, Annika; Beyersdorf, Friedhelm; Siepe, Matthias

Author Information
doi: 10.1097/MAT.0b013e3182523237
  • Free


The classic medical treatment options for extensive coronary artery disease (CAD) are percutaneous transluminal angioplasty or surgical treatment by bypassing coronary stenoses (coronary artery bypass grafting). Both treatment options aim to improve perfusion in the ischemic area. Those patients whose myocardial tissue has been destroyed or who are not amenable for conventional treatment remain challenging and their options for medical treatment are limited. Great effort is being taken to understand the mechanisms that lead to the human body’s ability to improve perfusion in areas of chronic ischemia. Different approaches are used to induce new vessel formation in the ischemic or damaged tissue as the first step of regeneration. As one of the first factors expressed at the onset of critical ischemia, vascular endothelial growth factor (VEGF) seems to play an important role in inducting new vessel formation. Strategies described to induce exogenous therapeutic neovascularization include growth factor transfer such as protein-or gene-based growth factor delivery, by viral encoded or naked plasmid DNA. Despite the great potential attributed to growth factor-induced therapeutic neovascularization, clinical trials results have been largely discouraging.1 In spite of the sobering results of growth factor transfer, recent trials emphasize both the mode of molecular delivery and the application mode.2 In our study we used skeletal myoblasts isolated and expanded from newborn Lewis rats. The cells were transfected with pCINeo-VEGF121.To avoid direct intramyocardial injection, the transfected cells were seeded on polyurethane (PU) scaffolds which served as the transport medium for cell transplantation.

Material and Methods

Isolation and Cell Culture of Primary Skeletal Myoblasts

Skeletal myoblasts (SkM) were isolated from male newborn Lewis rats as reported previously.3 In brief, limb muscles were minced and enzymatically digested with 1.5 mg/ml collagenase IA (Sigma Chemical Company, St. Louis, MO) and 0.25% trypsin/EDTA (Gibco, Auckland, New Zealand). Following passage through a 100-µm and a 40-µm filter (BD Falcon, Franklin Lakes, NJ), cells were expanded in culture a medium consisting of 57% Minimum Essential Medium (MEM) (Gibco), 27% M199 (Gibco), 15% fetal bovine serum (HyClone, Logan, UT), 1% penicillin/streptomycin (Gibco), and 5 ng/ml beta-FGF (Sigma). After 48 hours, cells were harvested by trypsinization, counted, and frozen in fetal bovine serum containing 10% dimethyl sulfoxide (Sigma) until they were needed for scaffold preparation. We usually obtained an average of 1.5 ± 0.7 × 107 cells per newborn Lewis rat. Antidesmin immunofluorescent staining (Sigma, clone DE-U-10) was performed to calculate the percentage of SkM in the preparation as described previously.4 Culture purity attained 46.8 ± 12.1% of desmin-positive myoblasts.

Preparation of Polyurethane Scaffolds and Cell Seeding

The polyurethane-urea matrix used in this study was obtained from Artimplant (Artelon, Västra Frölunda, Sweden). Artelon is a soft, elastic material which degrades by hydrolysis over approximately 5 years. The manufacture of Artelon membranes has been described.5 For our purpose, the biomaterial was processed into a highly porous scaffold (90 ± 3% volume porosity) with interconnected pores measuring 30–300 µm.

For scaffold seeding, Artelon membranes were cut to a scaffold size of 1 × 7.5 × 7.5 mm and precoated with 10 µg/ml laminin (Invitrogen, Carlsbad, CA) for 45 min at room temperature. SkM were thawed, diluted, centrifuged, and counted. Preliminary viability tests showed that cells were unaffected by the freeze-thaw process (data not shown). 30 µl of cell suspension containing 5.0 × 106 cells were applied to the scaffold. This amount of cell suspension disperses into the matrix and thereby distributes the cells to the polyurethane as shown in detail previously.6 The seeded scaffolds were incubated for 30 min at room temperature to allow cell attachment. Afterwards, culture medium was added and seeded PU scaffolds were incubated further at 37°C (5% CO2) for 48 hours. This seeding protocol was identified as the optimal way to facilitate cell attachment in our preceding study. We found that almost 50 % of cells attach.7 Following incubation, quantification of attached cells yielded 3.2 ± 0.6 × 106 cells in the scaffolds. For control, unseeded scaffolds were prepared in the same way using 30 µl of culture medium instead of cell suspension.

Transfection of Primary Skeletal Myoblasts

Cells were transfected as described previously.8 In brief, the plasmid pcDNA3-huda-Akt (expressing a myristinylated dominant-active form of Akt1) was delivered into SkM using the liposome-based transfection reagent Metafectene Pro (Biontex, Munich, Germany). According to the manufacturer’s protocol, we incubated 5 µg plasmid DNA (diluted in 100 µl MEM) with 12 µL Metafectene Pro solution (diluted in 100 µl MEM) for 25 min at room temperature. Following attachment of the cells, culture medium (2 ml) of the cell-seeded PU scaffolds was renewed and transfection mix was added. After 24 hours, the culture medium was changed again and the scaffolds incubated further at 37°C (5% CO2). The plasmid was a generous gift from Robert Freeman (University of Rochester, Rochester, NY). Effective transfection of the cells was confirmed by reverse transcriptase-polymerase chain reaction and Western Blot as reported in our previous study.8

Animal Model

The animals received humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the National Institute of Health (NIH Publication No. 86–23, revised 1985). All procedures were performed in accordance with the German Law on the Protection of Animals after obtaining permission from the Regional Administrative Authority (Freiburg, Germany, G-05/56).

The animal procedure has been described before.6 In brief, female Lewis rats were anesthetized with isoflurane (5% of oxygen for induction and 2.5% of oxygen for maintenance) and additive buprenorphine (10 µg/kg subcutaneously), placed on a warming pad (37°C), and tracheally ventilated at 80 cycles/min (14-gauge IV cannula, Abbocath, Abbott Ireland, Sligo, Republic of Ireland; Small Animal Ventilator 683, Harvard Apparatus Inc., Holliston, MA). A myocardial infarct was then induced through a left lateral thoracotomy by ligating the left anterior descending coronary artery. Two weeks later, a sternotomy allowed access to the infarction area for baseline catheterization and intervention. The animals were randomized to undergo a sham operation (group Sham, sternotomy and adhesiolysis only, n = 10), implantation of a scaffold seeded with transfected myoblasts (PU-VEGF-SkM, n = 9) (Figure 1A) or untreated myoblasts (PU-SkM, n = 9), or direct intramyocardial injection of transfected (Inj-VEGF-SkM, n = 10) or untreated (Inj-SkM, n = 9) myoblasts. In the PU-VEGF-SkM and PU-SkM groups, the scaffolds were placed on the epicardial surface of the visually identified infarction area and surgically fixated by four single epicardial sutures at the corners of the quadratic matrix (7/0 polypropylene). In the Inj-VEGF-SkM and Inj-SkM groups, 5.0 × 106 cells in 150 µl medium (containing 15% fetal bovine serum) were distributed within the scar via 4–6 intramyocardial injection sites. Six weeks after the intervention, the rats were anesthetized for left ventricular catheterization before the hearts were removed for histological analysis.

Figure 1
Figure 1:
Macroscopic and microscopic views of polyurethane-scaffold on the myocardial infarction zone (* indicates the scaffold, # indicates the viable myocardium).A: During final hemodynamical measrements severeal vessels are visible running towards the scaffold. The arrow points at the pressure-volume catheter inserted in the left ventricular apex. B: In an animal treated with unmodified skeletal myoblasts a large myocardial infarction area is visible just beneath the scaffold (Elastika van Gieson staining). C: The scaffold contains vascular endothelail growth factor–modified skeletal myoblasts. The ant-von Willebrand-staining reveals a number of vessels that are present in the neighboring zone of the scaffold. D: The scaffold containing unmodified skeletal myoblasts did not lead to any increased neovascularization in the scar region beneath the scaffold or the myocardial border zone.

Hemodynamic Assessment

Before therapeutic intervention and 6 weeks thereafter, left ventricular function was assessed by online left ventricle (LV) pressure-volume signals obtained by a 1.9–French pressure-volume catheter (FT212, Scisense Inc., London, Canada). After sternotomy, a purse-string suture was placed in the apical region of the left ventricle, where the catheter was inserted and placed along the LV long axis using a IV cannula (Abbocath) as a sheath. The catheter was connected to a Sigma-SA signal processor (CD Leycom, Zoetermeer, The Netherlands) for display and acquisition of pressure-volume loops. We registered 3 cycles of 20 beats under respiratory arrest to measure the LV function, quantified by heart rate, LV end-diastolic and end-systolic pressure, the maximum rate of LV pressure increase (dP/dt(max)) and decrease (dP/dt(min)), LV ejection fraction, and pressure half-time (PHT).

Furthermore, the isovolumic relaxation time constant (Tau) was determined with phase-plot analysis. In addition, we determined the isovolumic contraction (PISO) and, in single beat estimation, the end-systolic elastance (EES) and diastolic chamber stiffness constant (KED).

Data were transferred to the CIRCLAB data analysis software package, a generous gift from Dr. Paul Steendijk (University Medical Center, Leiden, The Netherlands).

Infarction Size and Angiogenesis

Hearts were explanted and sliced into three parts of equal thickness orthogonally to the heart axis. The pieces were frozen in Jung Tissue Freezing Medium (Leica Microsystems, Nussloch, Germany) and stored at −20°C. Cryosections of 5 µm were prepared and fixed with ice-cold acetone. An Elastika van Gieson staining kit (Merck, Darmstadt, Germany) was used to visualize the infarcted myocardium in each of the three slices (Figure 1B). Staining was performed according to the manufacturer’s instructions. The sections were then photographed with a digital camera. We calculated the living and infarcted areas, and then integrated the values of all sections to estimate the infarction size of the three slices using an image-analysis program (ImageTool2.0, Microsoft Corporation, Redmond, WA).

For anti-von Willebrand factor staining, endogenous peroxidase activity was quenched by peroxidase-blocking reagent (Dako Deutschland GmbH, Hamburg, Germany). The sections were then blocked with 10% goat serum and subsequently incubated with rabbit antihuman von Willebrand factor antibody (Dako, 1:250) for 1 hour at room temperature (Figure 1, C and D). All stainings were developed with HRP-goat antirabbit secondary antibody (Dako) followed by 3,3-diaminobenzidine tetrahydrochloride chromogen (Dako) and counterstained with hematoxylin (Merck). We counted the positively stained capillaries in 10 visual fields (magnification ×400) for each heart in the infarction zone, remote myocardium, and infarction border zone. Capillaries were identified by a one-layered vessel wall, a present or collapsed lumen, and a diameter of up to about 10 µm. Arterioles, arteries, and veins were excluded.

Statistical Analysis

Data are expressed as mean ± standard deviation. The statistical significance of differences between pre- and postinterventional data within one group was calculated using the Student’s t-test. One-way analysis of variance (ANOVA) was used to identify differences among the groups in capillary density and infarction size at the end of the investigation. Two-way ANOVA for repeated measurements was used to compare hemodynamic pre- and postinterventional variables among groups. Both ANOVA tests were followed by Holm-Sidak as a post hoc test. Differences were considered significant if p < 0.05.


A total of 47 female Lewis rats were randomized into the sham group (group Sham, sternotomy and adhesiolysis only, n = 10), the scaffold seeded with transfected myoblasts group, consisting of the treated and untreated myoblast subgroups (PU-VEGF-SkM, n = 9;PU-SkM, n = 9), or the direct intramyocardial injection group again consisting of two subgroups, namely the transfected (Inj-VEGF-SkM, n = 10) or untreated (Inj-SkM, n = 9) myoblasts. Figure 1 shows the size of the infarction zone in the different groups (Figure 2).

Figure 2
Figure 2:
Size of the infarction zone in percent of the whole heart. Three parts of the excised hearts were stained with Elastika van Gieseon staining to observe the infarcted myocardium. From the photographed heart slices, we calculated the living and infarcted areas using an image analysis program. Mean values ± standard deviation deviation are given. Sham, only sternotomy and adhesiolysis; PU-VEGF-SkM, scaffold seeded with transfected myoblasts; PU-SkM, scaffold seeded with untreated myoblasts; Inj-VEGF-SkM, direct intramyocardial injection of transfected myoblasts; Inj-SkM, direct intramyocardial injection of untreated myoblasts.

Figure 2 shows the capillary density in the infarction border zone in the different subgroups (Figure 3). The Inj-VEGF-SkM group’s blood vessel count was statistically significantly higher than the sham group’s (p < 0.05), and their capillary density was significantly higher than that of the PU-VEGF-SkM group (p < 0.01). Compared to the Inj-SkM group, the degree of capillary density in groups Inj-VEGF-SkM and PU-VEGF-SkM was statistically significant higher (p < 0.05). Table 1 summarizes the different capillary densities in the different parts of the myocardium in the various subgroups with the corresponding p-values.

Table 1
Table 1:
Capillary Density in the Different Parts of the Myocardium Representing Mean Counts of Identified Capillaries Per Visual Field
Figure 3
Figure 3:
Capillary density in the infarction border zone in each group. Values are given in percent of sham group which was normalized to 100%. Sham, only sternotomy and adhesiolysis; PU-VEGF-SkM, scaffold seeded with transfected myoblasts; PU-SkM, scaffold seeded with untreated myoblasts; Inj-VEGF-SkM, direct intramyocardial injection of transfected myoblasts; Inj-SkM, direct intramyocardial injection of untreated myoblasts.

Figure 4, A and B compare the hemodynamic data in the subgroups. LV ejection fraction (LVEF) in group PU-VEGF-SkM showed no significant increase after initiating angiogenesis by VEGF stimulation (pre 0.18 ± 0.06 vs. post 0.16 ± 0.11; p = 0.651). The Inj-VEGF-SkM group’s data is similar (pre 0.24 ± 0.12 vs. post 0.19 ± 0.7; p = 0.140). Maximum rate of LV pressure increase ((dP/dt(max) and decrease (dP/dt(min)) in the PU-VEGF-SkM group did not change significantly either (dP/dt(max) pre 2,,,157 ± 1,222 vs. post 2,135 ± 850; p = 0.963; dP/dt (min) pre −2,010 ± 1,251 vs. post −2,258 ± 1,106; p = 0.662). Analogically not significant data was collected for the Inj-VEGF-SkM (dP/dt(max) pre 1,905 ± 596 vs. post 2,499 ± 539; p = 0.133; dP/dt (min) pre −1,682 ± 596 vs. post −2,779 ± 2,136; p = 0.132). The only significant measured value was PHT in the PU-VEGF-SkM group (pre 9.7 ± 2.4 vs. post 13.9 ± 3.8; p = 0.004).

Figure 4
Figure 4:
Systolic (A) and diastolic (B) hemodynamic parameters in percent of the values before therapeutic intervention. LVEF, left ventricle ejection fraction; PISO, isovolumic contraction; Sham, only sternotomy and adhesiolysis; Inj-SkM, direct intramyocardial injection of untreated myoblasts; PU-SkM, scaffold seeded with untreated myoblasts; Inj-VEGF, injected with VEGF-overexpressing myoblasts; PU-VEGF, scaffold seeded with transfected myoblasts, PHT, pressure half-time; SB-ked, single beat diastolic chamber stiffness constant; SB-Emax, single beat maximal Elastance; dP/dt(max), maximum rate of left ventricle pressure increase; dP/dt(min), maximum rate of left ventricle pressure decrease.

We determined Tau via phase-plot analysis. Furthermore, we measured PISO and, in single beat estimation, EES, and the KED. None of these data revealed any significant changes comparing pre- and postinterventional values.

Despite having verified enhanced angiogenesis, measured hemodynamic values failed to show evidence of benefit in the VEGF-treated subgroups. Our sham group showed the biggest decreases in nearly all our hemodynamic parameters compared to the various subgroups, confirming the efficacy of our randomization strategy.


Investigation of the mechanism to induce neovascularization by proangiogenic gene therapy has led to different strategies involving growth factors and their transfer to the tissue. Vascular endothelial growth factor is one of the main growth factors in neovascularization playing a critical role in the angiogenetic response. Because the intravascular administration of protein does not seem appropriate for inducing adequate revascularization,1,9–11 we used plasmid DNA as a vehicle for VEGF transfer in our trial. To prevent tissue damage by needle injection, we used epicardially implanted cell-seeded polyurethane scaffolds, which proved to be a useful tool to study paracrine stem-cell action.6–8

In our study, we succeeded in demonstrating enhanced capillary density by transplanting VEGF-overexpressing skeletal myoblasts in the infarcted heart. Both groups with VEGF-overexpressing myoblasts (injection and scaffold-based) reveal a significant change in capillary density, with the scaffold group showing slightly greater density. Recent trials recommend gene transfer-based strategies as the most promising approach.1,12 In addition to the mode of gene transfer, the means of delivering infected myoblasts seems critical. One common strategy is to deliver the plasmid-infected myoblasts via injection. A limitation of this delivery strategy may be low transplantation efficiency because of uneven allocation in the target area and a high cellular death rate. In our trial, we employed the scaffold-based transfer of genetically modified cells to reduce the cellular death rate while aiming to achieve a more even allocation of the VEGF-modified myoblasts. This scaffold-based delivery strategy proved to be as efficient as the injection strategy in inducing angiogenesis.7 The impaired purity of the myoblast culture of almost 50% is in line with most small animal studies using these cells. The purity of myoblasts was not checked in the course of culture time for the current analysis. However, in our previous studies on the scaffold seeding optimization7 and the transfection optimization,8 we did not observe decreasing myoblast purity in the course of time. Furthermore, it is likely that the remaining cells in culture, the fibroblasts, will have a similar impactin terms of transfection and VEGF release. It has been shown recently, that the transfection of fibroblasts lasts even longer than for myoblasts.13 Nevertheless, the same purity was used for all myoblasts groups and the results of the intergroup comparison should be valid even with reduced purity.

We observed enhanced capillary density in both groups with VEGF-transfected myoblasts. We assumed that the gene transfer of VEGF by plasmid DNA leads via the paracrine effect to enhanced capillary density. Although the difference between the two VEGF groups in capillary density was not statistically significant (PU-VEGF-SkM versus Inj-VEGF-SkM), we favor the scaffold-based method because its less invasive character and even cell allocation.

Interestingly, we did not observe improvement in the VEGF-treated heart’s hemodynamics. Only the PU-VEGF-SkM group revealed significant change, namely in PHT. We assume that the formation of new capillary vessels in the chronically ischemic region is not strictly accompanied by myocardial tissue regeneration. One explanation for this observation may be that although we succeeded in inducing neovascularization, we did not achieve adequate amounts of therapeutic agent to produce a measurable hemodynamic benefit. This explanation is likewise one of our study’s limitations, because although we demonstrated VEGF overexpression of the transfected myoblasts, we can only speculate as to how much of the paracrine effect of VEGF actually reached the target cells. In addition, recent studies assume that the effectiveness of VEGF 121 seems to be lower in regard to formation of myotubes and free cells compared to VEGF 165, which we did not use in our study.

Another potential explanation is that the induction of enhanced capillary density does not suffice to restore ischemically damaged myocardium, suggesting that a more sophisticated mechanism may lead to angiomyogenesis for cardiac repair.

Lekas et al. analyzed various clinical trials evaluating protein and gene therapy, emphasizing that the intravascular administration of protein is insufficient to induce therapeutic angiogenesis, favoring gene transfer-based strategies. Moreover, their comparison of these clinical trials led them to conclude that neither protein- nor gene-based therapy provided evidence of therapeutic effectiveness in patients with coronary artery disease (CAD) or peripheral arterial disease.1 They assume that pro-arteriogenic strategies may be more successful if growth factors are administered to tissue sites containing nascent collaterals. Thus, they clearly differentiate between less effective angiogenesis and more effective arteriogenesis,in other words, enhancement of the formation of more mature collateral vessels. The double–blind, placebo-controlled clinical trial of Stewart investigated 93 CAD patients who received either VEGF plasmid DNA delivered via an endocardial route using a guided catheter, or placebo.12 Similar to Lekas’s findings, Stewart’s questions the therapeutic effect of intramyocardial plasmid VEGF gene therapy. Haider et al. investigated the potential of human skeletal myoblasts carrying human VEGF165 for angiomyogenesis for cardiac repair already in 2004. In a swine heart model, he demonstrated significantly improved regional blood flow and a rise in left ventricular ejection fraction 6 weeks after injection of the infected cells.14 We agree with Haider that combining gene therapy and cell transplantation leads to a measurable angiogenic response in hypoperfused myocardium. That same group published similar data in 2007 transplanting nanoparticle-transfected skeletal myoblasts in the rat heart model.2 Nevertheless, in our preclinical animal model, we failed to verify concurrent functional myocardial repair leading to improved cardiac performance as did Haider and Ye.

Lekas’s argument may help explain the discrepancy in our study, namely that the hemodynamics did not improve despite the demonstrated benefit regarding vessel formation in the VEGF groups after gene transfer via SkM. According to Lekas, this discrepancy may be because of the fact that we induced less effective angiogenenesis in our hearts to enable more effective arteriogenesis, which probably leads to angiomyogenesis and ultimately cardiac tissue repair. However, although this may explain why neovascularization did not result in a hemodynamic benefit in our study, we only demonstrated the correlation between the transplantation of VEGF-overexpressing skeletal myoblasts and capillary density in the ischemic heart which may be because of genuine neovascularization or simply the protective effect of existing capillaries in the control groups, or a combination of the two effects. Furthermore, we realized an incomplete transfection of the transplanted cells,8 which might contribute to the missing hemodynamical consequences6

In summary, the transplantation of VEGF-overexpressing skeletal myoblasts leads to enhanced capillary density in infarcted hearts.

The scaffold-based transfer of genetically modified cells proved to be as efficient as the injection strategy in inducing angiogenesis, and it remains a useful tool with which to study paracrine stem-cell action. However, we observed no hemodynamic benefit using VEGF overexpression.


1. Lekas M, Lekas P, Latter DA, Kutryk MB, Stewart DJ. Growth factor-induced therapeutic neovascularization for ischaemic vascular disease: time for a re-evaluation? Curr Opin Cardiol. 2006;21:376–384
2. Ye L, Haider HKh, Tan R, et al. Transplantation of nanoparticle transfected skeletal myoblasts overexpressing vascular endothelial growth factor-165 for cardiac repair. Circulation. 2007;116(11 Suppl):I113–I120
3. Leobon B, Garcin I, Menasche P, Vilquin JT, Audinat E, Charpak S. Myoblasts transplanted into rat infarcted myocardium are functionally isolated from their host. Proc Natl Acad Sci USA. 2003;100:7808–7811
4. Pouzet B, Vilquin JT, Hagège AA, et al. Intramyocardial transplantation of autologous myoblasts: can tissue processing be optimized? Circulation. 2000;102(19 Suppl 3):III210–III215
5. Gisselfalt K, Edberg B, Flodin P. Synthesis and properties of degradable poly(urethane urea)s to be used for ligament reconstructions. Biomacromolecules. 2002;3:951–958
6. Siepe M, Giraud MN, Pavlovic M, et al. Myoblast-seeded biodegradable scaffolds to prevent post-myocardial infarction evolution toward heart failure. J Thorac Cardiovasc Surg. 2006;132:124–131
7. Siepe M, Giraud MN, Liljensten E, et al. Construction of skeletal myoblast-based polyurethane scaffolds for myocardial repair. Artif Organs. 2007;31:425–433
8. Blumenthal B, Golsong P, Poppe A, et al. Polyurethane scaffolds seeded with genetically engineered skeletal myoblasts: a promising tool to regenerate myocardial function. Artif Organs. 2010;34:E46–E54
9. Henry TD, Annex BH, McKendall GR, et al. The VIVA trial: Vascular endothelial growth factor in Ischemia for Vascular Angiogenesis. Circulation. 2003;107:1359–1365
10. Simons M, Annex BH, Laham RJ, et al. Pharmacological treatment of coronary artery disease with recombinant fibroblast growth factor-2: double-blind, randomized, controlled clinical trial. Circulation. 2002;105:788–793
11. Lederman RJ, Mendelsohn FO, Anderson RD, et al. Therapeutic angiogenesis with recombinant fibroblast growth factor-2 for intermittent claudication (the TRAFFIC study): a randomised trial. Lancet. 2002;359:2053–2058
12. Stewart DJ, Kutryk MJ, Fitchett D, et al. VEGF gene therapy fails to improve perfusion of ischemic myocardium in patients with advanced coronary disease: results of the NORTHERN trial. Mol Ther. 2009;17:1109–1115
13. Gonçalves GA, Vassallo PF, dos Santos L, et al. Intramyocardial transplantation of fibroblasts expressing vascular endothelial growth factor attenuates cardiac dysfunction. Gene Ther. 2010;17:305–314
14. Haider HKh, Ye L, Jiang S, et al. Angiomyogenesis for cardiac repair using human myoblasts as carriers of human vascular endothelial growth factor. J Mol Med. 2004;82:539–549
Copyright © 2012 by the American Society for Artificial Internal Organs