Author + information
- Received February 5, 2015
- Revision received June 10, 2015
- Accepted June 12, 2015
- Published online August 25, 2015.
- Farhan Zafar, MD∗∗ (, )
- Robert B. Hinton, MD†,
- Ryan A. Moore, MD†,
- R. Scott Baker, BS∗,
- Roosevelt Bryant III, MD∗,
- Daria A. Narmoneva, PhD‡,
- Michael D. Taylor, MD, PhD† and
- David L. Morales, MD∗
- ∗Division of Pediatric Cardiothoracic Surgery, The Heart Institute, Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Cincinnati, Ohio
- †Division of Cardiology, The Heart Institute, Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Cincinnati, Ohio
- ‡Department of Biomedical Engineering, University of Cincinnati, Cincinnati, Ohio
- ↵∗Reprint requests and correspondence:
Dr. Farhan Zafar, Cincinnati Children’s Hospital, 3333 Burnet Avenue, ML 2013, Cincinnati, Ohio 45229.
Background Prosthetic valves currently used in children lack the ability to grow with the patient and often require multiple reoperations. Small intestinal submucosa-derived extracellular matrix (SIS-ECM) has been used successfully as a patch for repair in various tissues, including vessels, valves, and myocardium.
Objectives This study sought to assess the remodeling potential of a tubular tricuspid valve (TV) bioprosthesis made of SIS-ECM by evaluating its growth, structure, and function in a growing ovine model.
Methods A total of 12 3-month-old lambs were studied for a period of 3 or 8 months. SIS-ECM TVs were placed in 8 lambs; conventional bioprosthetic valves and native valves (NV) were studied as controls. All lambs underwent serial echocardiography, measuring annulus diameter and valve and right ventricular function.
Results The SIS-ECM valves demonstrated an incremental increase in annular diameter similar to NV. SIS-ECM valve function was normal in 7 of 8; 1 valve had severe regurgitation due to a flail leaflet. Explanted SIS-ECM valves approximated native tissue in gross appearance. Histopathology demonstrated migration of resident mesenchymal cells into the scaffold and trilaminar ECM organization similar to an NV, without inflammation or calcification at 8 months. Ex vivo mechanical testing of SIS-ECM valve tissue showed normalization of the elastic modulus by 8 months.
Conclusions In an ovine model, tubular SIS-ECM TV bioprostheses demonstrate “growth” and a cell-matrix structure similar to mature NVs while maintaining normal valve function. The SIS-ECM valve may provide a novel solution for TV replacement in children and adults.
Valvular heart disease represents a major public health problem globally, and is associated with significant morbidity and mortality (1,2). Each year, approximately 290,000 heart valve replacements are performed worldwide; this number is expected to increase 3-fold by 2050 (3,4). Additionally, many aggressive valve repair procedures are done each year because no long-lasting valve replacement options exist, especially in the pediatric population. The 2 main valve replacement options, mechanical and bioprosthetic (BP), improve survival and quality of life but possess significant and distinct limitations (5,6). Mechanical valves are prone to thromboembolic events and increased risk of serious hemorrhage due to required anticoagulation therapy. Because bioprostheses are more susceptible to structural valve degeneration, they lack durability, resulting in necessary reoperations. All of these potential adverse events are magnified in the pediatric population, where the need for a functional valve is decades longer than in the typical adult receiving valve replacement. Valve replacement in pediatric patients is further confounded by the need for small prosthetic valve sizes and, consequently, multiple operations to accommodate patient growth.
Small intestinal submucosa-derived extracellular matrix (SIS-ECM) scaffolds are capable of growth and remodeling into surrounding tissues such as the superior vena cava (7). SIS-ECM has been successfully used as a patch material in a wide variety of tissues, including bladder walls and ligaments, as well as some cardiovascular tissues, including vessels, myocardium, and to a limited extent, valves (8–17). In a small number of animals, acellular SIS-ECM–derived tubular valves have demonstrated early physiological remodeling (16), but the potential for growth and durable function remains unknown. Therefore, we hypothesized that a SIS-ECM tubular valve in the tricuspid position has the potential to remodel, as demonstrated by the organization of a cell matrix architecture similar to native valve (NV) tissue, as well as by normal growth and in vivo function over time. The objective of this study was to assess a tubular tricuspid valve (TV) bioprosthesis made of SIS-ECM by evaluating its growth, cell matrix architecture, and function in a growing ovine model.
Twelve lambs (mixed breed, 6 wethers and 6 ewes), approximately 3 months of age, were included in the study (Figure 1). SIS-ECM valves (20 mm) were implanted in 8 lambs, conventional BP valves in 2 lambs, and NVs in 2 lambs. For tissue collection and examination, 6 lambs were euthanized at 3 months and the remainder at 8 months. All lambs received humane care in compliance with The Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (18). All study protocols were approved by the Cincinnati Children’s Research Foundation Institutional Animal Care and Use Committee.
Valve material and design
SIS-ECM (CorMatrix Cardiovascular, Roswell, Georgia) is a sheet of acellular ECM derived from porcine small intestinal submucosa. Two of the SIS-ECM ply sheets are press-lyophilized together and shaped into a desired tubular structure by suturing its edges together with 6-O polypropylene suture (Prolene, Ethicon US, Somerville, New Jersey), resulting in a tubular valve with an inner diameter of 20 mm and a length of 30 mm (Figure 2). SIS-ECM is made up of approximately 90% collagen, the majority of which is type 1 collagen that is reported to maintain 3-dimensional architecture. It also contains glycosaminoglycans, including heparin sulfate, chondroitin sulfate, and hyaluronic acid (19).
The tubular SIS-ECM valve works similar to native atrioventricular valves. Native tricuspid and mitral valves develop in the form of a tube with dissociation from the inner lining of the embryonic cardiac tube and form chordae as they develop. External pressure as a result of ventricular contraction on the SIS-ECM tube collapses the tube just like a Heimlich valve. The 3 papillary attachment points create the 3 leaflets and there are no distinct leaflets in the tube (Online Video 1).
Surgical procedures and imaging
Lambs were intubated under general anesthesia using ketamine (5 to 15 mg/kg) and midazolam (0.02 to 0.05 mg/kg) for induction and isoflurane (1% to 4%) for maintenance; vecuronium (4 to 8 μg/kg) was used to induce paralysis as needed. A dose of ceftiofur sodium (1 to 2.2 mg/kg) was given as antibiotic prophylaxis and buprenorphine (0.005 to 0.01 mg/kg) for analgesia before incision. A right lateral thoracotomy was performed in the fourth intercostal space, and the pericardium was opened anterior to the phrenic nerve in a longitudinal fashion. Heparin (300 U/kg) was administered and re-dosed based on activated clotting time before cardiopulmonary bypass. After establishing cardiopulmonary bypass via aortobicaval cannulation and a cardioplegia (lidocaine–magnesium; single dose) arrest of the heart, a right atriotomy was performed. The TV apparatus, including 3 leaflets and all chordae tendineae, was excised. The distal end of the SIS-ECM valve was sutured to the 3 papillary muscles using a 5-O polydioxanone or 5-O polypropylene suture with a small supporting SIS-ECM pledget, which was introduced with the evolving experience after the first 2 lambs had already undergone the procedure. The first 2 lambs, therefore, had no pledgets used for the papillary attachments, but it was felt that the suture cut through the papillary muscle because of the repeated stress placed on the muscle/suture interface when the valve was functioning. Therefore, a SIS-ECM pledget subsequently was added to distribute the force on the tissue, a common surgical method used to address such a concern.
The proximal end of the SIS-ECM valve was sutured to the TV annulus in a running fashion, using the same suture material. Valve competence was tested by intraoperative saline injection leak test, and once confirmed, the right atriotomy was closed and cardiopulmonary bypass discontinued. A right thoracotomy was closed with a drain in place, and the lamb was allowed to recover in a cage. Lambs received supplemental oxygen via cone mask until vital signs were stable and they could stand up.
All lambs underwent transthoracic echocardiographic evaluation in the post-operative period, at 1.5 and 3 months, and for the second cohort, at 5 and 8 months following surgery. The reader was blinded to the valve type being evaluated for all echocardiography studies. Valve size (annulus dimension) and leaflet mobility (full, restricted, or flail) were noted. Valve function was assessed using established guidelines and valve dysfunction (regurgitation or stenosis) was classified as none, trace, mild, moderate, or severe (20,21). Right ventricular function was noted at each interval and classified as normal, mildly, moderately, or severely depressed (22).
An extended Methods section is available in the Online Appendix, including tissue harvesting, preparation, analysis, and biomechanical testing.
Echocardiographic measurements are presented as mean ± SD. Continuous data were compared using the Mann-Whitney U nonparametric test with Bonferroni corrections for multiple comparisons.
All lambs undergoing surgery recovered without complication and survived in good health until their scheduled euthanasia dates. There were no reoperations, bleeding, or wound infections noted in any animals, nor were there any paravalvular leaks or valve failures noted on post-operative echocardiogram in any lamb. All lambs, including the native controls, demonstrated normal activity, appetite, and weight gain throughout the study period.
SIS-ECM valve changes over time
Serial transthoracic echocardiography demonstrated physiological enlargement in the SIS-ECM valves as shown by an incremental increase in the annulus diameters proportional to the change in weight similar to NVs (Central Illustration). There was no change in the annulus diameter in BP valves. Seven of 8 SIS-ECM valves had stable trivial-to-mild regurgitation (Online Video 2). One SIS-ECM valve had severe regurgitation due to leaflet detachment from the papillary muscle. This valve had been attached to the papillary muscles without a pledget-reinforced suture (Online Video 3).
All SIS-ECM valves had full mobility without stenosis (transannular mean gradient 2.0 ±1.0 mm Hg), and right ventricular systolic function remained normal (Tables 1 and 2). Stable trace “physiological” tricuspid regurgitation was also noted for all BP and NVs. The BP valves developed stenosis by 8 months, as evidenced by an increasing transannular mean gradient (4.9 ± 1.6 mm Hg), as the sheep grew and gained weight, which was significantly higher compared with SIS-ECM valves (p = 0.004). The mean transannular gradient for the BP valve animal was 7.5 mm Hg at 8 months with restricted leaflet mobility (Table 2). Similarly, transannular velocity for SIS-ECM valves (0.68 ± 0.18 m/s) was similar to NVs (0.48 ± 0.15 m/s; p = 0.06); however, it was lower compared with the BP valve (1.1 ± 0.16 m/s; p = 0.004).
At 8 months, SIS-ECM valves had a similar appearance to NVs: grossly, they looked as if there were 3 leaflets attached to papillary muscles by rudimentary chordae tendineae (Figure 3). The margins of the SIS-ECM valve were integrated into the surrounding tissue at both the annulus and papillary muscles. There was no evidence of gross thrombosis or calcification. One SIS-ECM valve had a papillary detachment noted at the 1.5-month echocardiogram (Figure 4). This failure was attributed to a technical failure secondary to not using a SIS-ECM pledgetted suture to attach the valve to the papillary muscle.
Composite images of the entire TV leaflets show the SIS-ECM valve to be largely acellular with unorganized ECM after 3 months, but densely cellular with stratified ECM approximating normal architecture after 8 months (Figure 5). This is evidenced by separation of collagen, elastin, and proteoglycan layers restricted to the leaflet’s proximal, nonapposing aspect (Figure 5). Interestingly, the proximal aspect of the 3-month SIS-ECM valve shows migration of resident smooth muscle actin (SMA)-positive valve interstitial cells (VICs) into the scaffold with elongated cell morphology during constructive remodeling, but a return to normal VIC morphology after 8 months, suggesting transition from remodeling to homeostasis (23,24). Of note, fibroblast morphology at 3 months by ultrastructure electron microscopy was abnormal, demonstrating elongated spindles that normalized by 8 months, during which time collagen fibers increased and organized (Figure 6). Qualitatively, the vast majority of cells were resident cells that migrated in from the annular margin. After 3 months, there was partial endothelialization of the SIS-ECM valve demonstrated by CD31 staining. At 8 months, this endothelialization was complete on the valve surface and there was evidence of ongoing remodeling in that interstitial cells stained positive for CD31. There was nominal evidence of inflammation as demonstrated by small clusters of macrophages at both 3 and 8 months post-implantation (Figure 6), mostly around the suture areas, suggesting an injury mechanism with little evidence of infiltration of the macrophages by hematoxylin and eosin staining (data not shown). Fetal liver kinase 1 (vascular endothelial growth factor receptor 2; FLK1 [VEGFR2]) demonstrated increased expression in the interstitium, especially surrounding neovessels. Likewise, there was only trace mineralization, without gross calcification, at 3 months and none at 8 months (Figure 6).
Among NVs, the elastic Young’s modulus was low at 3 months after implantation (6 months of age) (Figure 7), as compared with the value for the Young’s modulus of a normal adult valve (Figure 7, dashed line). The elastic modulus of the SIS-ECM valves was similar to the normal adult valve value at both time points. Tissue stiffness was similar in the proximal and distal regions in both SIS-ECM and NVs at both 3 months and 8 months (data not shown). These results demonstrate that at 8 months post-implantation (11 months of age), the SIS-ECM valves have both physiological valve biomechanical properties and normal mature ECM architecture (Figures 5 and 6).
Tubular SIS-ECM valves increase in size proportional to the sheep’s somatic growth, which was similar to the normal growth of NVs in control animals. By the end of the study, sheep gained more than 3 times their original weight, their valve annulus showed a 50% increase in size (average 10 mm), and the valve remained physiologically functional (Central Illustration). Histologically, SIS-ECM valves demonstrated scaffold population with resident VICs and ongoing matrix remodeling, as evidenced by cellular morphology and ECM organization approximating normal valves over time. The tubular SIS-ECM valve has the potential to be a durable valve replacement option for patients with valvular heart diseases. Currently, severe valvular heart diseases lack a pharmacological treatment, and surgery remains the only effective management, but even that approach remains suboptimal (25). Despite significant improvements in prosthetic valves and surgical techniques, these have not proven to be a cure. Instead, NV disease is often traded for prosthetic valve disease. Nakano et al. (26) reported a 10-year risk of 30% for valve-related events after TV replacements. This outcome worsens for pediatric patients, with survival of only 57% and freedom from repeat valve replacement of 25% at 20 years following aortic valve replacements (27). Taken together, current valve replacement options are inadequate for treating valvular heart diseases, particularly in children.
A significant limitation to currently available valve replacements is their failure to accommodate somatic growth in the pediatric population, and as a result, these valves undergo nonstructural valve failure (28,29). Surgeons have oversized these valves in an attempt to accommodate for this problem; however, patient-to-valve size mismatch has been shown repeatedly to have worse outcomes (30,31). Conventional BP valves in this study demonstrated no change in annular size and increasing transvalvular gradient with increase in weight of the lambs eventually resulting in stenosis and immobile leaflets, which is consistent with clinical studies (32,33). The tubular SIS-ECM valves have demonstrated growth, therefore accommodating change in annulus diameter (>50% increase) with a more than 3-fold increase in weight over the study period. This characteristic is vital for pediatric patients because patient growth guarantees their need for re-replacement with current prosthetic valves. With no anticoagulation needed, tubular SIS-ECM valves may also be a valid option for adults, particularly young adults wishing to avoid reoperations. Growth here refers to the change in annulus diameter with the change in animal weight; however, whether this is the same proliferative process as in an NV is unclear. Long-term studies are currently underway in our lab to further evaluate cellular proliferation and matrix changes in SIS-ECM valves to see whether true growth can be demonstrated.
Calcification is the most common reason of BP valve failure (34). Saleeb et al. (35) demonstrated only an 18% freedom from valve failure after 3 years of aortic valve replacement. In this experience, a bovine pericardial BP valve was used, and they demonstrated accelerated degeneration with diffuse intrinsic leaflet calcification with virtually immobile leaflets (35). We have noted only minimal calcification at 3 months and none at 8 months, with all valve leaflets fully mobile at last examination. Similarly, a few small areas of inflammation were noted, mostly around suture areas, which is a known problem with polypropylene suture. This is surprising because the juvenile ovine model (≤20 weeks at implant) is used to analyze calcification predilection in BP valves. In this model, such calcification is seen as early as 3 weeks post-implant, and in some valves, calcium content increases within days of implant; the average time to significant calcification of BP valves is 3 months post-implant (36,37). The SIS-ECM valve appears to be less prone to calcific degeneration than commonly seen in currently available BP valves.
This study demonstrated that the expression of alpha-SMA (alpha-SMA) was similar when comparing NVs with 3- and 8-month SIS-ECM valves, suggesting alpha-SMA-positive VICs are not activated as part of a disease process, and the favorable remodeling process we observed in the SIS-ECM valves is not mediated by the alpha-SMA VIC population. SIS-ECM demonstrated partial endothelialization of the scaffold with a small population of CD31-positive endothelial cells invading the interstitium at 3 months and more so by 8 months. This phenomenon resembles a nonpathological epithelial–mesenchymal transition–like process seen in embryonic (pre-natal) development, where endothelial cells acquire the migratory characteristics of mesenchymal cells and form organized tissue during development (38). This is also indicated by presence of interstitial CD31 cells and neovessel formation (VEGFR2-positive cells), consistent with processes resembling fetal wound healing (39). Taken together, this suggests that reactivated EMT may contribute to the remarkable post-natal remodeling that occurs in this SIS-ECM BP tissue. Further studies are needed to establish underlying mechanisms.
SIS-ECM valves are stiffer at 3 months compared with the NV, but demonstrate physiological stiffness similar to NV stiffness by 8 months, corresponding temporally with the changes in ECM architecture of the NV during growth. Valve elasticity changes over time in humans, pigs, and sheep, because they have more elastin fibers in fetal life and early age; however, with aging, elastin fibers are surrounded by more collagen fibers, reducing elasticity and increasing stiffness (40,41). The distal half of the valve leaflets lacks organization, which is consistent with progressive remodeling. Because there is no disease in these valves, we do not think this represents pathology. This observation is consistent with a lack of remodeling in the more distal aspect of the valve (i.e., with the remodeling process originating at the attachment point to native tissue and working its way from annulus to tip). Alternatively, this observation may reflect different regions of the valve experiencing different biomechanical stresses because the free edge of the valve experiences less shear. Normal in vivo (echo) and ex vivo (biomechanics) tissue function at 8 months post-implant suggests that the remodeling processes appear to be adaptive and durable. Further, ECM organization in the SIS-ECM BP is promising, but mechanisms of ongoing tissue homeostasis and potential for structural degeneration need to be examined more closely. This is being evaluated by long-term (≥2 years) SIS-ECM TV implants in the ovine model, which are already in the post-implant observational stage in our lab.
Due to using a large animal model, this study involved a relatively small number of animals. Therefore, some of the observations are limited to qualitative inferences, rather than statistical and quantitative conclusions. Leaflet lengths were not compared because this information was not consistent across the NVs due to the presence of well-formed chordae. Although the tissue analysis showed various valve-like cells and macrostructure, with no significant inflammation component apparent, the presence of various nonresident cells was not examined. Another important limitation of this study is the difference in coagulation cascades of ovine species from humans. Though there was no post-cardiopulmonary bypass anticoagulation used in these animals and no thrombotic events or valvular thrombosis noted, these results need to be cautiously interpreted.
Despite these limitations, this study demonstrated that SIS-ECM valves are capable of remodeling and growth, with intact valve function, in a rapidly growing animal model. Further studies to delineate the origin of these cells into the ECM, signals that modulate this remodeling, and valve examination over an extended period of time will help improve design and applications of the SIS-ECM valves. There are several unpublished reports of clinical use of SIS-ECM valves in young adults outside and on compassionate use basis inside the United States. Furthermore, phase I human studies are being considered by the U.S. Food and Drug Administration to examine the safety of these SIS-ECM bio-prostheses in the tricuspid position.
SIS-ECM valves demonstrated a potential to “grow” with the animal while maintaining normal valve function. The SIS-ECM valve also developed a cellular and matrix structure, as well as mechanical properties similar to the native ovine valve, which is unique for a BP valve demonstrating no calcification. The SIS-ECM valve bioprosthesis offers the potential to alleviate some of the limitations of current valve therapies by perhaps providing a novel solution for TV replacement in children and adults.
COMPETENCY IN MEDICAL KNOWLEDGE: The long-term success of valve replacement in patients with valvular heart disease is limited by structural degeneration of biological prostheses or need for continuous anticoagulation with mechanical devices, and when implanted in young patients, somatic growth typically requires reoperation. Tubular valves formed from SIS-ECM have shown growth in an animal model.
TRANSLATIONAL OUTLOOK: Further studies are needed to delineate the origin of the cells populating the ECM, understand the signals that modulate its remodeling, and evaluate the properties of valves created from SIS-ECM with respect to capacity for continued growth, and resistance to thrombosis, calcification, and structural degeneration over extended periods.
The authors thank Dr. Naritaka Kimura (surgeon for several valve implantation procedures), Amy Opoka (histopathologic analysis of valve tissue), and Hanna Osinska (electron microscopy of valve tissue) for their assistance.
For an expanded Methods section and supplemental videos, please see the online version of this article.
This project was funded by the Cincinnati Children's Hospital. Dr. Morales serves on the medical advisory board of CorMatrix Cardiovascular without compensation; is a clinical events committee member of Berlin Heart; is a Total Artificial Heart proctor for Syncardia; has received research funding from CorMatrix; and has received fees for travel and accommodations from CorMatrix. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- extracellular matrix
- native valve
- small intestinal submucosa-derived extracellular matrix
- tricuspid valve
- valve interstitial cell
- Received February 5, 2015.
- Revision received June 10, 2015.
- Accepted June 12, 2015.
- American College of Cardiology Foundation
- Mozaffarian D.,
- Benjamin E.J.,
- Go A.S.,
- et al.
- Hammermeister K.,
- Sethi G.K.,
- Henderson W.G.,
- Grover F.L.,
- Oprian C.,
- Rahimtoola S.H.
- Schoen F.J.
- Zhao Z.Q.,
- Puskas J.D.,
- Xu D.,
- et al.
- Gilbert C.L.,
- Gnanapragasam J.,
- Benhaggen R.,
- Novick W.M.
- National Research Council (US) Committee for the Update of the Guide for the Care and Use of Laboratory Animals
- Rudski L.G.,
- Lai W.W.,
- Afilalo J.,
- et al.
- Hinton R.B. Jr..,
- Lincoln J.,
- Deutsch G.H.,
- et al.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Une D.,
- Ruel M.,
- David T.E.
- Friedman S.,
- Edmunds L.H. Jr..,
- Cuaso C.C.
- Saleeb S.F.,
- Newburger J.W.,
- Geva T.,
- et al.
- Quinn R.W.
- Nistico P.,
- Bissell M.J.,
- Radisky D.C.