Congenital cardiovascular defects remain the leading cause of infant death from congenital malformations.1 Congenital heart surgeons use many different materials to repair or close myocardial and vascular defects or to create tunnels and conduits. Currently, no ideal patch material exists. The materials available for use, including polyethylene terephthalate (PET;Dacron), expanded polytetrafluoroethylene (ePTFE;GORE-TEX) and membrane materials such as bovine pericardium, are associated with significant morbidity. These materials can produce significant inflammatory responses leading to calcification and contraction over short periods of time, as well as infection or thrombosis. They do not remodel or degrade and thus cannot grow with the pediatric patient, often necessitating reoperation later in life.2 Even native materials such as autologous pericardium, which do not trigger an immune response, will over time fibrose, retract, and can become aneurysmal.3
A clinical need exists for a bio-resorbable material for use in cardiac repair applications. One approach is the use of decellularized extracellular matrix scaffolds (ECM).4 In contrast to synthetic biodegradable materials, decellularized scaffolds contain ECM that is proven to be chemoattractant, re-cellularize, and remodel.5–7 Once implanted, ECM is populated with a variety of cell types from adjacent tissues as well as marrow-derived stem cells from the circulation, making it capable of growth, integrating into surrounding tissues, and maintaining regional viability.5,8,9 Following migration and attachment, cells, some of which may be stem cells, can undergo proliferation, differentiation, and phenotypic maturation, remodeling the implanted scaffold into functional tissue.4,10 Furthermore, because nearly all of the porcine cells are removed from the scaffold during processing, and because collagen is the primary matrix element remaining (> 90%) and well-conserved amongst species, ECM materials are considered minimally immunogenic.11,12
In preclinical and clinical use, ECM-derived scaffolds have demonstrated neovascularization, remodeling, and restoration of regional function in various tissues.6,7,13,14 These materials are used clinically for musculotendinous replacement, bladder reconstruction, dura mater replacement, and body wall repair. A commercial extracellular matrix product has also been approved in the United States and Europe for human cardiac repair and pericardial closure but does not include an indication for remodeling in the United States.14 In preclinical models, this material has been shown to repopulate with myocytes7,11,15,16 and to confer regional mechanical benefit in canine, ovine and bovine cardiac-repair models, leading to certification mark approval for cardiac tissue repair in Europe with an indication for remodeling.17 Recent work demonstrated that tricuspid valve replacement with a tubular ECM valve in a growing lamb remodels into functional valvular tissue resembling native tissue.6,18 The current study assessed longitudinal electrical, mechanical and histologic properties of a commercially available ECM patch following replacement of a large area of right-ventricular (RV) wall in juvenile sheep.
Materials and Methods
The study protocol was approved by the Institutional Animal Care and Use Committee and all animals received humane care in compliance with “The Guide for the Care and Use of Laboratory Animals,” published by the US National Institutes of Health. Eleven juvenile Suffolk-mix sheep (five wether males, six females) at 4 months of age (32 ± 4 kg) received an ECM patch (CorMatrix SIS-ECM, CorMatrix Cardiovascular Inc, Atlanta, GA). Every 3 months sheep underwent cardiac magnetic resonance imaging (MRI) and three-dimensional-electrophysiologic mapping. Sheep were euthanized for histological assessment as follows, two at 3 months, three each at 6, 9, and 12 months.
CorMatrix SIS-ECM is derived from decellularized sheets of porcine small intestinal submucosa. Four sheets of non-crosslinked SIS-ECM are press-lyophilized together to create the patch material and then sterilized with ethylene oxide gas. The ECM patch was soaked in saline for 30 seconds prior to implant.
All lambs were fasted 12 hours and received preoperative antibiotic and analgesia. Under general anesthesia a right 4th intercostal space lateral thoracotomy was performed and lambs placed on cardiopulmonary bypass.6,18 A 2 × 2 cm square, full-thickness defect in the Right ventricular outflow tract (RVOT) wall was excised, and a 2 × 2 cm ECM patch was sewn to the defect’s border using running 5-0 polydioxanone suture. Each patch corner was marked with 4 mm radiopaque clips on the epicardial side only (Figure 1A). Lambs were fully recovered before moving to the housing suite.
Cardiac MRI Procedure
Under general anesthesia, cardiac MRI was performed (1.5T, Ingenia, Philips Healthcare, The Netherlands) every 3 months after patch placement. Standard steady-state free precession (SSFP) images were obtained with retrospective electrocardiographic gating while freely breathing in standard planes. Late gadolinium enhancement (LGE) imaging was performed 10 minutes after contrast injection using standard inversion recovery gradient echo sequence with inversion time chosen to null the left ventricular myocardium.
Ventricular Volumes and Strain Analysis
Myocardial strain was assessed from RV short stack SSFP cine images using a feature tracking algorithm (Tomtec, Diogenes MRI, Germany).19,20 Among three directional strains, (circumferential, radial, longitudinal) RV circumferential strain statistically correlates with RV function using feature tracking with cardiac MRI19,20 and speckle tracking echocardiography.21 Therefore, we compared average circumferential strain of the patched region to RV free wall, excluding the patched region and septum.
Contractile function of the patch and surrounding tissue was determined using strain assessment and local deformational parameters (multidirectional strain) calculated. Late gadolinium enhancement images were measured in the myocardium and patch and assessed qualitatively and quantitatively (normalized signal area).
Three-Dimensional Electrical Mapping
Every three months under general anesthesia, venous access was obtained and a 11 French deflectable tip sheath was placed in the right atrium and a seven French CARTO NogaStar mapping catheter was advanced through the sheath and fluoroscopic guidance used to place the tip on each radiopaque clip to generate a three-dimensional anatomic, voltage, and activation map in the patch area and surrounding areas during sinus and then ventricular paced rhythm using the CARTO XP mapping system (Biosense Webster). Ventricular pacing was performed from three locations in different directions from the patch and activation and voltage maps repeated.
Calculation of Conduction Velocities
Velocities were calculated using custom software built using Matlab (Mathworks, Natick, MA). Three-dimensional pacing, electrode, and patch vertex coordinates, voltages, and time differentials were read directly from the CARTO output file for each pacing location. Euclidean distances between all location points and patch vertices and the reference point were calculated. Location points were assigned to being within or outside the patch based on the predefined patch vertices using a standard ray-casting algorithm. Velocities were calculated for points within the patch and those outside the patch. Patch velocities were calculated from the front edge of the patch to the measured point. Myocardial velocities were calculated from the reference point to the measured point.
ECM patch cell populations were compared with native RV myocardial cells from the same hearts. Samples were collected and analyzed from the center of the patch to the patch periphery. To identify tissue composition, Masson’s trichrome, Movat’s pentachrome, and Alizarin red stains were used.22–24 For identification of cell types immunofluorescence on paraffin sections was used. Primary antibodies: sarcomeric myosin MF-20 (Hybridoma Bank, IA), connexin43 (Invitrogen, CA), smooth muscle actin, vimentin (Sigma, MO), CD45 (Serotec, NC) were visualized with corresponding secondary antibodies conjugated with Alexa488 or 568. Nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI)-(Molecular Probes, OR). Images were acquired using Nikon A1 laser confocal microscope system.
Data are presented as mean and standard deviation with significance set at α = 0.05. There are multiple functional measurements per sheep, so we accounted for sheep-to-sheep variation by 2 way analysis of variance. Individual measurements from 3 month intervals are comparable to prior measurements by paired t-test and all sheep via independent sample t-test. Conduction velocities and voltages are represented as means and error bars represents 95% confidence intervals; comparisons were made using independent sample t-test for each time point.
All sheep grew normally and remained healthy throughout the protocol. Gross visual assessment of the ECM patches at 3, 6, 9, and 12 months was suggestive of integration with surrounding tissues with noticeable patch thickening by 9 and 12 months (Figure 1B–F). At necropsy, there was gross evidence of infarct in three ECM specimens (not shown), likely due to transection of coronary branches when removing myocardium at implantation.
The LGE images consistently showed a hyper-intense area surrounded by normal myocardium representing the ECM patch (Figure 2A), which did not change in appearance over the 12 month study. The local circumferential strain was initially positive indicating a noncontractile dyskinetic segment (Figure 2B, C). By 9 months, regional strain became negative indicating a trend toward normalization of regional contractility. By 12 months, regional contractility had normalized and was not different than remote RV regions of the same heart. Noncontractile areas of ECM patch diameter shrank over time, when comparing the 3 month to 9 month scans, (Figure 3A–C). There was an increase in viable myocardium extending in from edges of the patch.
Three-Dimensional Electrical Mapping
Serial electrophysiologic mapping conducted at the 3, 6, 9, and 12 month timepoints consistently demonstrated conduction through and across the patch and no animal demonstrated reentrant arrhythmia. There was no difference in electrogram voltage between the patch and adjacent myocardium with the exception of a slight reduction in patch voltage at 9 months only (Figure 4A). Conduction velocities were initially slower in the patch compared to the adjacent myocardium, but were similar at 6 and 12 months (Figure 4B). In three animals an area of myocardium adjacent to the patch had low voltage consistent with infarct injury noted at necropsy.
At 3 months postsurgery, Masson’s trichrome staining provided benchmark morphology in the normal sheep ventricle (Figure 5A). In surgically modified hearts at 3 months, both Masson’s trichrome staining (Figure 5B–D) and Movat’s pentachrome staining (Figure 5E–F) indicate the nature and extent of fibrosis. The interior of the patch area already had been vascularized (Figure 5B). The patch area and proximal areas to it were devoid of intact muscle tissue, although some patches of muscle could be identified in some sections (Figure 5D, F), but were abnormal, lacking the usual tightly bundled architecture of normal muscle (Figure 5A). Abundant fibrosis was also present in and around the patch with both collagen and fibrin apparent (Figure 5E). Elastin is abundant in areas containing adipocytes (Figure 5F). In normal ventricle, well-organized cardiomyocytes with defined intercalated disks containing connexin 43, a marker of gap junctions, are present (Figure 6A). At this initial, 3 month time point, in contrast to normal heart, the interior patch area is devoid of cardiomyocytes and intercalated disks, consistent with absence of cardiac muscle (Figure 6B). There is, however, a sparse but significant and reproducible population of puncta staining positively for connexin 43 (Figure 6B, insert), which is consistent with the ability of fibroblasts and smooth muscle to passively conduct current,25 as consistently demonstrated via three-dimensional electrical mapping of the ECM patch region. Immunohistochemical analysis using CD45 antibody clearly showed the presence of lymphocytic infiltration (Figure 6C), but over time those areas appear to decrease in size and number (not shown). Consistent with vasculature in the patch (Figure 5B), numerous cells stained positively for vimentin and smooth muscle actin, protein markers for vascular smooth muscle (Figure 6D, E). The later time points (6, 9, and 12 months) did not differ materially from these initial data, except that we noted less inflammatory response after 6 months, (See Figure 1 to 3, Supplemental Digital Content, http://links.lww.com/ASAIO/A320). Focal areas of calcification were seen around suture material and minimally in one ECM sample at 12 months; all other ECM patches were devoid of calcification.
The present investigation was designed to test ECM patch durability and viability under conditions that would typically exceed those of most clinical applications such as a transannular patch used for repair of Tetralogy of Fallot. We evaluated the ability of SIS-ECM to patch a large, full thickness RV defect and to assess its potential to either support function or partially remodel into functional myocardium in a rapidly growing lamb model. There were no patch failures or postoperative mortality associated with the ECM patch, despite the radical excision of myocardium required to initially create the 2 × 2 cm RV defect. However, we did observe areas of apparent infarct adjacent to three patches, likely reflecting the disruption of coronary blood supply in creating the large defect in a relatively small RV (juvenile sheep). This may have attenuated the current study results compared to the remodeling seen in previous studies where a different, potentially less traumatic defect was created.7,15,17,26 Despite this, all eleven ECM patches remained viable throughout their individual 3, 6, 9, 12 month endpoints.
Magnetic resonance imaging showed the ECM patch was initially noncontractile with displacement in the opposite direction of surrounding myocardium at 3 and 6 months (as seen with aneurysm), but trended towards normal (shortening) at 9 months and was not different from the native myocardium by 12 months. We did not observe any detrimental changes in other metrics of global or regional RV function. A recent study also used cardiovascular magnetic resonance to compare ECM and Dacron full thickness implants in a 2 month long porcine model of RVOT replacement. They showed superior elasticity and perfusion with ECM versus Dacron at 2 months; however all measures remained inferior to normal myocardium, which is expected given the shorter study duration.27 These findings are also consistent with prior studies using measurement of Young’s modulus to show that ECM implant stiffness can transform over time to match that of native myocardium or valvular architecture.6,7 Importantly, ECM adaptively remodels to meet variable requirements of different cardiac environments, acting as a scaffold for contractile cells in myocardial applications and maintaining stiffness when used for tricuspid valve replacement.6
Serial electrophysiologic mapping showed consistent conduction through and across the ECM patch which at 6 months and again at 12 months was similar to conduction through adjacent tissue by paced velocity measurement suggesting it appeared to normalize with that of native myocardium and furthermore no animals demonstrated reentrant arrhythmia. This is an encouraging result for ECM use considering the large myocardial defect and that ECM is initially nonconductive. A prior study also used the CartoXP mapping system to characterize electromechanical properties of urinary bladder-derived ECM full thickness implants after 2 months in porcine model of RVOT replacement.28 They observed low-level electrical activity in the patch that increased when moving toward the margins, which is consistent with our observations that demonstrated progressive increase in voltage and conduction velocity over time. These results are consistent with a previous study demonstrating the ECM patch is capable of transducing coordinated electrical signals at 5 and 8 months after implant using a hanging heart model.7 The signal frequency observed in the center of the patch also matched the pacing stimulation frequency, indicating the patch contains electrically excitable cells. The immunohistochemistry presented in the current study demonstrates that connexin 43 is expressed and this, along with prior similar studies, is consistent with the hypothesis that the ECM patch can remodel from an inert scaffold into electrically conductive tissue.7,28
Histological examination of ECM patches and adjacent myocardium showed the ECM patch was consistently populated with viable cells that consisted mostly of fibroblasts, smooth muscle cells, adipocytes and endothelial cells. Noticeably absent in the patch were cardiomyocytes, with only a few abnormal bundles present in the periphery. The only significant change in cell population noted was attenuation of the inflammatory response after 6 months. There was minimal to no calcification observed except immediately adjacent to suture material. A recent study also demonstrated an absence of calcification at 18 months follow-up in a growing sheep model that used ECM for tricuspid valve replacement.18 This is notable because other synthetic and bioprosthetic implants in juvenile sheep are prone to calcification, often evident within a few weeks and with complete calcification in months.29 Moreover, nonexistent or very minimal calcification has been noted in surrounding tissue of pediatric cardiac ECM explants,30 or during imaging studies of long-term patients.31
Histology from 2 month studies in pigs after a 2 × 2 cm full-thickness RV defect substituted with ECM patch indicated the presence of nascent myocardium on the endocardial surface, substantial angiogenesis and alpha-smooth muscle actin-positive cells,27,28 similar to what we observed histologically at 3 months in this sheep study. However, importantly we did not see replacement of the ECM with native cardiac cells by 12 months, though measures of RV wall function and electrical characteristics continued to improve. The absence of cardiac cells or myocardium was also noted in a small group of samples from a pediatric clinical study in which cardiac ECM patches were explanted, yet ECM viability and function were clinically good.30 Of note, it has also been shown histologically that ECM is not totally decellularized at implant and therefore some remnant cells could be interpreted as indicators of neovascularization or myocardial remodeling as observed in many short-term studies.7,17,26–28
The SIS-ECM material has the potential to be a durable, viable cardiovascular tissue replacement for congenital heart surgery patients. The ECM patch appears to exceed the functional capabilities and remodeling abilities of bovine pericardium, Dacron or Gore-Tex. Furthermore, ECM may eventually serve as a stable matrix for attracting cytokines and other signaling molecules that support and steer cellular populations to potentially enhance function or contractile potential; abilities pericardium and Gore-Tex have failed to show. Preliminary reports of cardiac ECM performance in clinical applications are encouraging,31,32 indicating low complication (< 5%) and reoperation rates (2%), but have noted inflammation is sometimes associated with long-term, explanted ECM material from humans.30 Early degeneration of ECM and aneurysm formation have also been reported and are important aspects to consider, particularly when ECM is used for repairs in higher pressure environments like aortic arch repair.33 However, it should be noted that because only a small percentage of clinical ECM implants performed are ever explanted, thus the vast majority may actually exhibit a more encouraging degree of remodeling not documented histologically. A longer follow-up is needed to understand the real potential and contraindications of this material, particularly in pediatric applications where time will define its growth potential.
The limitations of this study include the small sample size and although statistics were applicable for comparing changes in functional parameters, data from the smaller sized histologic groups are merely observational, and only suggest the potential for ECM patch remodeling. Perhaps the model disadvantaged the amount of cellular remodeling as seen in other models by the excision of such a relatively large amount of the right ventricle that clearly showed evidence of peri-patch ischemia not noted in previous studies.
In a growing ovine model, an ECM patch becomes viable, functional tissue and remains so up to a year, while demonstrating some aspects of remodeling as evidenced by mechanical, electrical and histologic properties. However, the ECM was devoid of cardiomyocytes and populated instead with vascular, collagen, elastin, adipose, and fibrotic cells. None-the-less this material presents a potentially better solution for cardiac grafting and tissue repair, especially in the pediatric population where it might prevent future reoperation. More clinically specific animal models are needed to further confirm these findings, followed by clinical studies.
1. Roger VL, Go AS, Lloyd-Jones DM, et al.; American Heart Association Statistics Committee and Stroke Statistics Subcommittee: Heart disease and stroke statistics—2011 update: A report from the American Heart Association. Circulation 2011.123: e18–e209.
2. Oechslin EN, Harrison DA, Harris L, et al. Reoperation in adults with repair of tetralogy of fallot: Indications and outcomes. J Thorac Cardiovasc Surg 1999.118: 245–251.
3. Kumar SP, Prabhakar G, Kumar M, et al. Comparison of fresh and glutaraldehyde-treated autologous stented pericardium as pulmonary valve replacement. J Card Surg 1995.10: 545–551.
4. Badylak SF, Freytes DO, Gilbert TW. Extracellular matrix as a biological scaffold material: Structure and function. Acta Biomater 2009.5: 1–13.
5. Reing JE, Zhang L, Myers-Irvin J, et al. Degradation products of extracellular matrix affect cell migration and proliferation. Tissue Eng Part A 2009.15: 605–614.
6. Zafar F, Hinton RB, Moore RA, et al. Physiological growth, remodeling
potential, and preserved function of a novel bioprosthetic tricuspid valve: Tubular bioprosthesis made of small intestinal submucosa-derived extracellular matrix. J Am Coll Cardiol 2015.66: 877–888.
7. Scully BB, Fan C, Grigoryan B, et al. Remodeling
patch into functional myocardium
in an ovine model: A pilot study. J Biomed Mater Res B Appl Biomater 2016.104: 1713–1720.
8. Badylak SF, Park K, Peppas N, McCabe G, Yoder M. Marrow-derived cells populate scaffolds composed of xenogeneic extracellular matrix. Exp Hematol 2001.29: 1310–1318.
9. Zhao ZQ, Puskas JD, Xu D, et al. Improvement in cardiac function with small intestine extracellular matrix is associated with recruitment of C-kit cells, myofibroblasts, and macrophages after myocardial infarction. J Am Coll Cardiol 2010.55: 1250–1261.
10. Badylak SF. The extracellular matrix as a scaffold for tissue reconstruction. Semin Cell Dev Biol 2002.13: 377–383.
11. Badylak S, Obermiller J, Geddes L, Matheny R. Extracellular matrix for myocardial repair. Heart Surg Forum 2003.6: E20–E26.
12. Badylak S, Kokini K, Tullius B, Simmons-Byrd A, Morff R. Morphologic study of small intestinal submucosa as a body wall repair device. J Surg Res 2002.103: 190–202.
13. Lopes MF, Cabrita A, Ilharco J, et al. Esophageal replacement in rat using porcine intestinal submucosa as a patch or a tube-shaped graft. Dis Esophagus 2006.19: 254–259.
14. Scholl FG, Boucek MM, Chan KC, Valdes-Cruz L, Perryman R. Preliminary experience with cardiac reconstruction using decellularized porcine extracellular matrix scaffold: Human applications in congenital heart disease. World J Pediatr Congenit Heart Surg 2010.1: 132–136.
15. Potapova IA, Doronin SV, Kelly DJ, et al. Enhanced recovery of mechanical function in the canine heart by seeding an extracellular matrix patch with mesenchymal stem cells committed to a cardiac lineage. Am J Physiol Heart Circ Physiol 2008.295: H2257–H2263.
16. Tanaka A, Kawaji K, Patel AR, et al. In situ constructive myocardial remodeling
of extracellular matrix patch enhanced with controlled growth factor release. J Thorac Cardiovasc Surg 2015.150: 1280–90.e2.
17. Kochupura PV, Azeloglu EU, Kelly DJ, et al. Tissue-engineered myocardial patch derived from extracellular matrix provides regional mechanical function. Circulation 2005.112(9 Suppl): I144–I149.
18. Baker RS ZF, Moore R, Taylor MD, Morales DL. Tubular bioprosthetic tricuspid valve implant demonstrates chordae formation and no calcification—long-term follow-up. J Am Coll Cardiol. 2017;70:2456–2458.
19. Heermann P, Hedderich DM, Paul M, et al. Biventricular myocardial strain analysis in patients with arrhythmogenic right ventricular cardiomyopathy (ARVC) using cardiovascular magnetic resonance feature tracking. J Cardiovasc Magn Reson 2014.16: 75.
20. Schuster A, Kutty S, Padiyath A, et al. Cardiovascular magnetic resonance myocardial feature tracking detects quantitative wall motion during dobutamine stress. J Cardiovasc Magn Reson 2011.13: 58.
21. Hayabuchi Y, Sakata M, Kagami S. Right ventricular myocardial deformation patterns in children with congenital heart disease associated with right ventricular pressure overload. Eur Heart J Cardiovasc Imag. 2015.16: 890–899.
22. Hinton RB Jr, Lincoln J, Deutsch GH, et al. Extracellular matrix remodeling
and organization in developing and diseased aortic valves. Circ Res 2006.98: 1431–1438.
23. Krenz M, Gulick J, Osinska HE, Colbert MC, Molkentin JD, Robbins J. Role of ERK1/2 signaling in congenital valve malformations in Noonan syndrome. Proc Natl Acad Sci U S A 2008.105: 18930–18935.
24. Ricmond G. Clearing and staining of embryo for demonstrating ossification. Stain Technology. 1938;13:77–79.
25. Kohl P, Gourdie RG. Fibroblast-myocyte electrotonic coupling: Does it occur in native cardiac tissue? J Mol Cell Cardiol 2014.70: 37–46.
26. Kelly DJ, Rosen AB, Schuldt AJ, et al. Increased myocyte content and mechanical function within a tissue-engineered myocardial patch following implantation. Tissue Eng Part A 2009.15: 2189–2201.
27. Tanaka A, Kawaji K, Patel AR, Ota T. The extracellular matrix patch implanted in the right ventricle evaluated with cardiovascular magnetic resonance protocol to assess regional physio-mechanical properties. Interact Cardiovasc Thorac Surg 2017.24: 82–89.
28. Ota T, Gilbert TW, Badylak SF, Schwartzman D, Zenati MA. Electromechanical characterization of a tissue-engineered myocardial patch derived from extracellular matrix. J Thorac Cardiovasc Surg 2007.133: 979–985.
29. Lehr EJ, Rayat GR, Chiu B, et al. Decellularization reduces immunogenicity of sheep pulmonary artery vascular patches. J Thorac Cardiovasc Surg 2011.141: 1056–1062.
30. Woo JS, Fishbein MC, Reemtsen B. Histologic examination of decellularized porcine intestinal submucosa extracellular matrix (CorMatrix) in pediatric congenital heart surgery
. Cardiovasc Pathol 2016.25: 12–17.
31. Quarti A, Nardone S, Colaneri M, Santoro G, Pozzi M. Preliminary experience in the use of an extracellular matrix to repair congenital heart diseases. Interact Cardiovasc Thorac Surg 2011.13: 569–572.
32. Witt RG, Raff G, Van Gundy J, Rodgers-Ohlau M, Si MS. Short-term experience of porcine small intestinal submucosa patches in paediatric cardiovascular surgery
. Eur J Cardiothorac Surg 2013.44: 72–76.
33. Erek E, Aydin S, Suzan D, Yildiz O, Demir IH, Odemis E. Early degeneration of extracellular matrix used for aortic reconstruction during the Norwood operation. Ann Thorac Surg 2016.101: 758–760.