EXPERIMENTAL Decellularized Tendon-Bone Composite Grafts for Extremity Reconstruction: An Experimental Study Simon Farnebo, M.D., Ph.D. Colin Y. L. Woon, M.D. Joel A. Bronstein, B.S. Taliah Schmitt, M.D. Derek P. Lindsey, M.S. Hung Pham, B.S. Alesha B. Castillo, Ph.D. James Chang, M.D. Stanford, Calif.
Background: Restoration of biomechanical strength following surgical reconstruction of tendon or ligament insertion tears is challenging because these injuries typically heal as fibrous scars. The authors hypothesize that injuries at the tendon-bone interface would benefit from reconstruction with decellularized composite tendon-bone grafts. Methods: Tendon-bone grafts were harvested from Sprague-Dawley rats. Grafts subjected to decellularization were compared histologically and biomechanically with untreated grafts ex vivo and in a new in vivo model. Wistar rats underwent Sprague-Dawley allograft reconstruction using a pair-matched design. The rats were killed at 2 or 4 weeks. B-cell and macrophage infiltration was determined using immunohistochemistry, and explants were tested biomechanically. Results: Decellularization resulted in a decrease in cells from 164 ± 61 (untreated graft) to 13 ± 7 cells per high-power field cells (p < 0.005) and a corresponding significant decrease in DNA content, and preserved scaffold architecture of the tendon-bone interface. Biomechanical comparison revealed no difference in failure load (p = 0.32), ultimate tensile stress (p = 0.76), or stiffness (p = 0.22) between decellularized grafts and untreated controls. Following in vivo reconstruction with tendon-bone interface grafts, decellularized grafts were stronger than untreated grafts at 2 weeks (p = 0.047) and at 4 weeks (p < 0.005). A persistent increase in B-cell and macrophage infiltration was observed in both the capsule surrounding the tendon-bone interface and the tendon substance in untreated controls. Conclusion: Decellularized tendon-bone grafts display better biomechanical properties at early healing time points and a decreased immune response compared with untreated grafts in vivo. (Plast. Reconstr. Surg. 133: 79, 2014.)
I
njuries to the site where tendons and ligaments insert onto bone are common and constitute a significant reconstructive challenge. Common examples include rupture of flexor tendon insertion at the distal phalanx,1 rupture of tendons in the rotator cuff,2,3 and rupture of the anterior cruciate ligament.4 Current reconstructive strategies involve hardware, suture anchors, and drill holes to approximate tendons and ligaments to From the Division of Plastic Surgery, Veterans Affairs Palo Alto Health Care System, and the Division of Plastic Surgery, Stanford University Medical Center. Received for publication April 17, 2012; accepted June 5, 2013. Presented at the 67th Annual Meeting of the American Society for Surgery of the Hand, in Chicago, Illinois, September 6 through 8, 2012. Copyright © 2013 by the American Society of Plastic Surgeons DOI: 10.1097/01.prs.0000436823.64827.a0
bone. However, the native tendon-bone interface is typically never fully reconstituted after surgery. Although gross healing may be present,5 microscopic inspection of the reconstructed tendonbone interface will reveal fibrous scar tissue with inferior biomechanical properties compared with the native insertion.6–9 Increased risk for failure ensues and, as a result, current surgical repair techniques have high reinjury rates. Up to 94 percent Disclosure: None of the authors has a financial interest in any of the products, devices, or drugs mentioned in this article.
This work was supported by the plastic surgery foundation.
www.PRSJournal.com
79
Plastic and Reconstructive Surgery • January 2014 of repaired massive rotator cuff tears fail,10 and 56 percent of anterior cruciate ligament reconstruction patients experience knee pain 1 year after surgery.11 Conceptually, the use of a decellularized tendon-bone extracellular matrix construct to replace a tendon-bone interface deficit is an attempt to replace the weaker tendon-to-bone healing with bone-to-bone and tendon-to-tendon healing, which is regarded as stronger because it involves the approximation of tissues with equivalent elastic moduli.12–14 Currently available extracellular matrix scaffolds approved for augmentation of tendon healing are primarily of animal origin.15,16 As with all nonhuman tissue grafts, concerns regarding the degree of decellularization and strong immune responses toward these xenogeneic materials have been raised.17–20 Human cadaveric allograft tissue represents an underused, widely available tissue source that may provide an alternative scaffold option for tissue-engineered composite grafts.21–25 Previous research on decellularized human flexor tendons26,27 and multiple composite tissue scaffolds (tendon-bone interface and ligament-bone grafts)28–30 may provide alternative treatment options in reconstructive surgery. These allografts contain extracellular matrix proteins such as collagen, elastin, and proteoglycans that are crucial for cell attachment, migration, and proliferation.31,32 It is believed that the decellularized extracellular matrix is insoluble and conserved across species and therefore does not trigger an immune response on reimplantation, compared with cellular debris, which is much more immunogenic.33,34 Therefore, it is important that allograft scaffolds used for upper extremity reconstruction be decellularized.24–26,35 A previous model of human flexor tendons has also shown that chemical decellularization reduces cellular material to a minimum (94 percent decrease of cellular content)26 and that this process eliminates the major histocompatibility complexes that are responsible for immune recognition33,34,36 and thereby reduces the immune response in vivo.29,35 The rat Achilles tendon-bone interface model was used to represent the flexor digitorum
profundus insertion at the distal phalanx. The rat Achilles tendon was chosen because (1) its superficial location enables easy access to surgical transection and subsequent reconstruction, (2) the rodent calcaneus is a bone that is large enough to tolerate clinical hardware (microdrill and screw), (3) it allows for comparison in pair-matched analysis with the contralateral side, and (4) it is of a similar order of magnitude (similar size ratio) as human digital flexors. In this article, we aim to show that decellularization, a necessary step for addressing immunogenicity and disease transmission, does not adversely affect scaffold integrity and strength properties after treatment. We hypothesize that (1) composite grafts can successfully be decellularized with preservation of biomechanical properties and structure, (2) decellularization substantially decreases the inflammatory response after in vivo allograft reconstruction, and (3) decellularized grafts are stronger at the early healing time points than untreated grafts in vivo.
MATERIALS AND METHODS Animals and Experimental Groups Eighty-six Sprague-Dawley (mean weight, 280 g; Charles River Laboratories, Willimantic, Conn.) Achilles tendon–calcaneus composite tendon-bone interface grafts were harvested. The grafts were divided into four groups and assessed with histology and biomechanical strength testing (Table 1). Grafts used as untreated allograft controls were snap-frozen in saline to −70°C after harvest and remained frozen until further use. Decellularization Targeted ultrasonication (VC505; Sonics & Materials, Newtown, Conn.) against the tendonbone interface of the grafts was performed in a chilled water bath (10 minutes; total, 64,800 J).29 Buffered 5% peracetic acid (Sigma, St. Louis, Mo.), 1% ethylenediaminetetraacetic acid, and 2% sodium dodecyl sulfate in 1% ethylenediaminetetraacetic acid was used for chemical
Table 1. Data for the Four Groups Group In vitro 1 2 In vivo 3 4
Description
Total No.
Histology
Strength
DNA Content
Untreated controls, SD grafts Decellularized grafts, SD grafts
19 23
5 9
9 9
5 5
Untreated controls, SD grafts in Wistar rats Decellularized grafts, SD grafts in Wistar rats
22 22
2 (2 wk) + 2 (4 wk) 2 (2 wk) + 2 (4 wk)
9 (2 wk) + 9 (4 wk) 9 (2 wk) + 9 (4 wk)
0 0
SD, Sprague-Dawley.
80
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts decellularization according to a modification of a previously established protocol.26 All grafts were rinsed with deionized water to ensure that grafts were chemical free before implantation. Our decellularization protocol was used as the sole sterilizing agent, as previous studies have shown that 5% peracetic acid and detergents efficiently minimize disease transmission.32,37 In Vitro Experiments Histologic Analysis Untreated and decellularized tendon-bone interface grafts were fixed in 10% formalin and decalcified overnight (Decal; Decal Chemical Corp., Tallman, N.Y.). Specimens were embedded in paraffin; sectioned (6 μm); and stained with hematoxylin and eosin, SYTOGreen fluorescent nucleic acid stain (Invitrogen, Carlsbad, Calif.), and Masson’s trichrome to determine the adequacy of decellularization and preservation of collagenous architecture. Cell Counts Using hematoxylin and eosin–stained sections, cell counts were performed by two independent observers. Eight random representative high-power field (20×) (Nikon TS100; Nikon, Melville, N.Y.) images were centered on the tendon-bone interface with a superscript grid, to ensure that the evaluated area examined remained the same in all samples. DNA Content Untreated and decellularized tendon-bone interface grafts were trimmed, lyophilized, and weighed. Total DNA was then isolated from each sample
(DNAeasy; Qiagen, Hilden, Germany). DNA content was calculated from the absorption at 280 nm analyzed in triplicate using a Spectrophotometer (ND-1000; NanoDrop, Wilmington, Del.) and normalized to the initial dry weight of the sample. Biomechanical Testing A pair-matched design was used to compare material properties between in vitro groups and between in vivo groups (untreated versus decellularized). Harvested grafts were secured in a custom-made clamp (Fig. 1) with the dissected foot locked securely under a metal bar and the tendon end pulled to failure (ultimate failure load) of the graft (MTS 858; MTS Systems Corp., Eden Prairie, Minn.). Orthogonal photographs together with a known scale (digital caliper) were obtained using a digital camera (resolution, 2592 × 1936 pixels) at a fixed distance to measure two dimensions of the tendon. The cross-sectional area was calculated using an elliptical assumption (ImageJ software; National Institutes of Health, Bethesda, Md.) and used to calculate the ultimate tensile stress (force divided by cross-sectional area). The constructs were pulled until failure at a rate of 0.5 mm/minute. The MTS system recorded force and displacement data continuously during each test. Stiffness was calculated using the gradient of the force-displacement curve during the linear phase of deformation. In Vivo Experiments Surgical Technique We used a rat allograft model (implantation of Sprague-Dawley allografts into Wistar rats; mean
Fig. 1. Biomechanical setup. (Left) Degloved rat foot with intramuscular portion of the Achilles tendon (1), Achilles tendon (2), and calcaneus bone (3). (Center and right) Photographs of rat foot in custom-made clamp (a) with the tibia locked securely outside the holding block (b) in two positions perpendicular to each other. The tendon ends were fastened between sheets of sandpaper using commercial cyanoacrylate glue and clamped between the crossheads of a servo-controlled hydraulic materials testing system (MTS 858). Arrows show direction of pull.
81
Plastic and Reconstructive Surgery • January 2014 weight, 320 g) to represent the scenario of clinical allograft reconstruction of the tendon-bone interface. The procedure is described in the legend for Figure 2. After surgery, all animals were allowed immediate mobilization without the need for a cast. All animal procedures were performed with an institutional animal care and use committee– approved animal protocol. Immunohistochemistry Tendon-bone interface grafts were harvested at 2 and 4 weeks for histologic analysis. For B-cell staining, a goat anti-rat CD20 primary antibody was used (Santa Cruz Biotechnology, Santa Cruz, Calif.). For macrophage staining, a mouse antirat CD68 primary antibody was used (Santa Cruz Biotechnology). Rat spleen was used as a positive control. Micro–Computed Tomography High-resolution (10.5-μm voxel size) postoperative imaging with micro–computed tomography (Viva CT 75; Scanco Medical AG, Brüttisellen, Switzerland) was performed for closer visualization of screw placement in the graft and calcaneus bone in four of the animals (Fig. 2). Statistical Analysis Data were reported as mean ± SD. Comparisons of continuous data across groups were performed
using paired t test. The significance level was set at p < 0.05. All analyses were performed using PASW Statistics for Windows, Version 18.0 (SPSS, Inc., Chicago, Ill.).
RESULTS Histology (In Vitro) Untreated controls demonstrated cellularity in tendon, fibrocartilage, and bone (Fig. 3, above, left, and center, left). Following decellularization, there was reduced cellularity in all parts of the constructs, including the fibrocartilage zone of the tendon-bone interface, as shown by hematoxylin and eosin– and Masson trichrome– stained images (Fig. 3, above, right, and center, right)) and SYTOGreen nucleic acid stain (Fig. 3, below, right). There were no visible changes in the morphology of the tendon, fibrocartilage, or bone following treatment, indicating microstructural preservation with almost complete removal of visible cells. Cell Counts (In Vitro) Physicochemical decellularization resulted in significant reduction of visible cells as shown with cell counts. Average cell counts decreased from 164 ± 61 cells/high-power field (untreated grafts)
Fig. 2. The Achilles tendon was sharply released from the calcaneus bone, and the posterior tuberosity of the calcaneus was dissected free. The bone was cut with an oscillating microsagittal saw approximately 3 mm distal to the tendon insertion (Stryker, Kalamazoo, Mich.) and thread holes were drilled (0.75 mm; Medartis AG, Basel, Switzerland) in the center of the bone. The graft bone segments were then drilled (0.75 mm; Medartis) and fixed with 0.9 mm (diameter) × 10 mm (length) self-tapping microscrews (Medartis) (left). The proximal tendon was cut proximally at the musculotendinous junction and sutured to the free end of the transected Achilles tendon using a 4-0 Prolene (Ethicon, Inc., Somerville, N.J.) locking stitch. Tendon-bone interface composite construct with 0.9-mm microscrew in place (left). Composite construct implanted in rat as an allograft (center). Micro–computed tomographic image of graft and screw in calcaneus bone (right).
82
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 3. Physicochemical decellularization resulted in significant reduction of visible cells and nucleic material, with microstructural preservation. Hematoxylin and eosin–stained untreated tendon-bone interface (above, left) and decellularized tendon-bone interface (above, right) (original magnification, × 10 for both). Masson trichrome–stained untreated tendon-bone interface (center, left) and decellularized tendon-bone interface (center, right) (original magnification, × 20 for both). Hematoxylin and eosin–stained decellularized tendon-bone interface (below, left) (original magnification, × 20) and SYTOgreen nuclear stain decellularized tendon-bone interface (below, right) (original magnification, × 20). T, tendon; FC, fibrocartilage; B, bone.
to 13 ± 7 cells/high-power field (decellularized grafts) (p = 0.0004). DNA Content The mean DNA content of the untreated grafts was 57.9 ± 7.5 ng/mg DNA/dry weight, compared with 21.7 ± 9.0 ng/mg DNA/dry weight, a significant decrease of 63 percent (p < 0.005). Biomechanical Testing (In Vitro and In Vivo) All biomechanical data are listed in Table 2, including mode of failure.
Animal Behavior (In Vivo) In the first week, the reconstructed animals showed a slight bilateral swelling around their feet. However, they started immediately to bear weight on both legs and walked around their cages with only a slight limp that disappeared after the first 2 weeks. Histology (In Vivo) At 2 weeks after implantation, increases in cellular infiltration and vascularization of the capsule were seen in both groups. A clear
83
Plastic and Reconstructive Surgery • January 2014 Table 2. Comparison of Data* Group In vitro 1 2 In vivo 3 4 3 4
Ultimate Failure Load (N)
Ultimate Tensile Stress (MPa)
Stiffness (N/mm)
Untreated controls Decellularized grafts
75.7 ± 18.5 69.5 ± 6.0 (NS)
10.2 ± 4.9 10.7 ± 3.0 (NS)
17.9 ± 5.5 15.2 ± 3.0 (NS)
2 wk, untreated grafts 2 wk, decellularized grafts 4 wk, untreated grafts 4 wk, decellularized grafts
23.3 ± 9.1 31.7 ± 7.4† 22.7 ± 9.6 46.9 ± 12.7‡
Description
1.2 ± 0.6 1.9 ± 0.6† 1.1 ± 0.5 2.3 ± 0.7‡
6.4 ± 2.0 7.1 ± 2.7 (NS) 4.3 ± 1.1 7.6 ± 0.9‡
Mode of Failure All tendon All tendon TBI, 5; tendon, 3; bone, 1 TBI, 4; tendon, 4; bone, 1 TBI, 6; tendon, 2; bone, 1 TBI, 7; tendon, 1; bone, 1
NS, nonsignificant (p > 0.05); TBI, tendon-bone interface. *Pair-matched samples: untreated grafts versus decellularized grafts in vitro, at 2 weeks, and at 4 weeks. In all cases of bone failure [n = 4 (11%)], a fracture of the drill canal through the graft bone was observed. An asymmetrical drilling of the graft canal was seen in three of these cases as a possible explanation; no slipping of the screws was seen in any of the grafts. †p < 0.05. ‡p < 0.005.
noninflammatory cellular infiltration of the tendon was seen only in decellularized grafts. Immunohistochemistry showed that macrophages were confined to the capsule and absent in the tendon substance, indicating the classic and sequential pattern of inflammatory cell accumulation as seen postoperatively after reconstruction with an allograft. Immunohistochemistry of B cells, in contrast, displayed a clear difference between the groups, with infiltration throughout both the capsule and the tendon substance in the untreated grafts (Figs. 4 and 5). This B-cell infiltration of the tendon, and the capsule, persisted and increased after 4 weeks, compared with the decellularized grafts. Also, macrophage infiltration of the capsule remained more prominent in the untreated grafts compared with the decellularized grafts (Fig. 6). The inflammatory cells appeared to invade the tendons in “sheets” physically separating the collagen fibers. Over time, more spindleshaped cells appeared in both groups, indicating a normal repopulation of the tissues.
DISCUSSION This article highlights the need for allograft decellularization to decrease graft immunogenicity and to accelerate postoperative healing in reconstructive surgery. Also, it demonstrates how decellularized composite tissue grafts offer an alternative reconstructive solution to complex tendon-bone ruptures in vivo. The use of decellularized tendon-bone constructs may thus challenge the existing paradigm of suture-based tendon-bone interface repair, suggesting that the damaged tendon-bone interface can be replaced with a biocompatible decellularized “off-the-shelf” tendon-bone interface construct. The four-zone tendon-bone interface transmits loads and decreases the concentration of
84
stress at the attachment zone.38 Collagen II–producing chondrocytes of the fibrocartilage are likely to be a key element in this biomechanical function38; however, they typically do not regenerate after injury, which precludes regeneration of the fibrocartilage zone.5,39 Therefore, previous studies have aimed at improving the healing of the fibrocartilage by inducing stress,40 local application of stem cells,41 use of growth factors,42 and use of fibrin glue.43 We have previously established a protocol to decellularize single tissue flexor tendon scaffolds using a combination of detergents26 and increasing microstructural scaffold porosity with peracetic acid.27 However, the fibrocartilage zone is only partially decellularized using these protocols.36 This is because the chondrocytes within the fibrocartilage are deeply embedded within their lacunae in a tight-knit, partially mineralized, stiff tendon fiber matrix. We have therefore added a mechanical component to the protocol to increase detergent penetration in the hard-to-penetrate fibrocartilage zone.29 Targeted ultrasonication as a first step is effective because it increases tissue porosity and mechanically disrupts cell membranes before extraction of nuclear material by the detergent solution.44 Microscopic analysis revealed a nearly complete reduction (92 percent) of cellular elements and nuclear material removed during the process of decellularization without collateral damage to the extracellular matrix structure. It has recently been shown in a flexor tendon xenograft model that decellularization is crucial to limiting an immune response in vivo and that a vigorous immune reaction can rapidly compromise biomechanical strength.35 An absent or much-attenuated immune response will be beneficial during initiation of early mobilization. The biomechanical
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 4. Hematoxylin and eosin staining (original magnification, × 20) showing tendon substance of untreated graft (above, left) and decellularized graft (above, right) at 2 weeks, and untreated graft (second row, left) and decellularized graft (second row, right) at 4 weeks. Although it is impossible to distinguish donor cell persistence in the untreated grafts (above, left and second row, left) from host cell infiltration, the greater number of cells in the untreated grafts is likely to be of inflammatory origin, based on their abundance seen only in the untreated grafts and also based on their morphology. Remaining tenocytes typically present as elongated cells (arrowheads, below, right) whereas the macrophages and lymphocytes are round (arrowheads, below, right). High-power (original magnification, × 40) insets of untreated graft at 2 weeks (third row, left) with
85
Plastic and Reconstructive Surgery • January 2014
Fig. 5. Immunohistochemistry staining (original magnification, × 20) of CD20+ lymphocytes in the tendinous part of an untreated graft (above, left) and decellularized graft (above, right) at 2 weeks, untreated graft (below, left), and decellularized graft (below, right) at 4 weeks.
strength of our decellularized construct is comparable to what was observed in previous studies on the Achilles tendon insertion in the calcaneus bone.41,45 Also, and more importantly, the ultimate failure load is comparable to the strength of a pullout suture model where healing was augmented by mesenchymal stem cells.41 However, future studies, using this model will be needed to determine its benefits compared with a conventional suture pullout repair and when this extracellular matrix scaffold is reseeded with pluripotent cells. The Sprague-Dawley–to-Wistar rat allograft model was intended to simulate clinical allogenic composite tissue transplantation. Because of species inbreeding, single-species rat isograft reconstruction resembles autograft replacement in design and may not sufficiently replicate the immune response expected in clinical allograft transplantation between patients. Reliable markers to evaluate the degree of immune response were chosen to Fig. 4. (Continued) polymorphonuclear infiltration in “sheaths” separating the tendon weave (asterisks), and surrounding vessels (third row, right). Untreated donor tendon (dT) at the region of tendon suture of the recipient rat Achilles tendon (rT) at 20× (below, left) and 40× (below, right) magnification illustrating the vast inflammatory response toward untreated graft tendon but not recipient tendon.
86
allow for adequate explanation of biomechanical testing results. CD20 and CD68 cell surface markers were used to identify B-cell and macrophage invasion at time points chosen to reflect the early inflammatory phase of postoperative healing. Our results concur with previous studies that have shown that macrophages are present during the early inflammatory phase of tendon and tendon-bone healing. These macrophages are known to produce proinflammatory mediators that may have a negative effect on tissue regeneration and strength, especially in the early “destructive” phase of inflammatory healing. Classically activated macrophages release proinflammatory cytokines such as interleukin-1, interluekin-6, and tumor necrosis factor-α, and matrix proteases, which aid in cell chemotaxis through the inflammatory milieu. It has been shown that tendon-bone healing is enhanced through reestablishment of collagen fiber continuity, new bone formation, and the overall degree of interface remodeling if macrophages are depleted from the injured zone.46 B-cell infiltration of the tendon core and capsule indicates that the humoral pathway of the immune system is also activated, likely leading to collagen degradation and compromised tendon structure. The status of grafts in this study resembles what we previously described
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 6. Immunohistochemistry (original magnification, × 20) staining of CD20+ lymphocytes at 4 weeks, showing an increased number of inflammatory cells in the capsule surrounding the tendon-bone interface in the untreated group (above, left) compared with the decellularized group (above, right). CD68+ macrophages at 4 weeks, showing an increased number of inflammatory cells in the capsule (c) and tendon (t) surrounding the tendon-bone interface in the untreated group (below, left) compared with the decellularized group (below, right).
in decellularized human flexor tendons used as xenografts in subcutaneous pockets of rats, where remaining major histocompatibility complexes in untreated tendons triggered a similar strong inflammatory response.35 Compared with untreated grafts, decellularized grafts were 1.4 times stronger at 2 weeks, and 2.0 times stronger at 4 weeks. These results can be explained by the weaker inflammatory response to decellularized grafts compared with untreated grafts, leading to superior biomechanical properties in the early healing period. Early strength will allow for initiation of early active motion protocols, facilitating earlier recovery of function and return to activities, and also impedes adhesion formation.40,47 A fine balance must be achieved between unloading and overloading of the tendon-bone interface. Earlier studies of conventional reattachment of the tendon to bone indicate that early mobilization is beneficial for extracellular matrix production at the repair site.48,49 However, excessive motion and load can also be detrimental to healing, as this extracellular matrix is composed of inferior scar-like material.50 Conservative postoperative protocols involving lengthy immobilization
are therefore often advocated. These protocols lead to gradual accumulation of tendon tensile strength at the expense of gliding. A closer analysis of pair-matched constructs revealed a significantly higher stiffness in the decellularized grafts at 4 weeks in vivo, compared with the untreated grafts. The stiffness of the untreated grafts continues to drop over time (from 6.4 N/mm to 4.3 N/mm) without a corresponding drop in ultimate failure load or ultimate tensile strength, whereas stiffness in the decellularized grafts improves only slightly (from 7.1 N/ mm to 7.6 N/mm) despite a 48 percent increase in ultimate failure load (from 31.7 N/mm to 46.9 N). The clinical impact of these discrepancies remains unknown, and we hesitate to draw conclusions for two reasons. First, we see a potential risk that when grafts are ruptured at various locations (e.g., tendon left and tendon-bone insertion right), the stress or stiffness would not be comparable. Second, stress values from composite tissue testing must be regarded with caution when composite tissue grafts are used, as the cross-sectional area at the side of insertion is very irregular, variable, and not necessarily elliptical.
87
Plastic and Reconstructive Surgery • January 2014 CONCLUSIONS Decellularized tendon-bone interface grafts possess ideal characteristics of an allograft scaffold—basic tendon and tendon-bone interface structure, native extracellular matrix architecture, and optimal biocompatibility.51 In the future, these grafts may be used to precisely match deficient tendon and complex tendon-bone insertion tears without donor-site morbidity.52 James Chang, M.D. Division of Plastic Surgery Stanford University Medical Center 770 Welch Road, Suite 400 Stanford, Calif. 94305 [email protected]
ACKNOWLEDGMENT
This project was funded by a Plastic Surgery Foundation grant, an American Foundation for Surgery of the Hand/American Society for Surgery of the Hand grant, and a Veterans Affairs Rehabilitation Research and Development Merit Review grant. REFERENCES 1. Moiemen NS, Elliot D. Primary flexor tendon repair in zone 1. J Hand Surg Br. 2000;25:78–84. 2. Lehman C, Cuomo F, Kummer FJ, Zuckerman JD. The incidence of full thickness rotator cuff tears in a large cadaveric population. Bull Hosp Jt Dis. 1995;54:30–31. 3. Yamaguchi K, Ditsios K, Middleton WD, Hildebolt CF, Galatz LM, Teefey SA. The demographic and morphological features of rotator cuff disease: A comparison of asymptomatic and symptomatic shoulders. J Bone Joint Surg Am. 2006;88:1699–1704. 4. Brown CH Jr, Carson EW. Revision anterior cruciate ligament surgery. Clin Sports Med. 1999;18:109–171. 5. Thomopoulos S, Genin GM, Galatz LM. The development and morphogenesis of the tendon-to-bone insertion: What development can teach us about healing. J Musculoskelet Neuronal Interact. 2010;10:35–45. 6. Liu CF, Aschbacher-Smith L, Barthelery NJ, Dyment N, Butler D, Wylie C. What we should know before using tissue engineering techniques to repair injured tendons: A developmental biology perspective. Tissue Eng Part B Rev. 2011;17:165–176. 7. Forward AD, Cowan RJ. Experimental suture of tendon to bone. Surg Forum 1960;11:458–460. 8. Galatz LM, Rothermich SY, Zaegel M, Silva MJ, Havlioglu N, Thomopoulos S. Delayed repair of tendon to bone injuries leads to decreased biomechanical properties and bone loss. J Orthop Res. 2005;23:1441–1447. 9. Newsham-West R, Nicholson H, Walton M, Milburn P. Long-term morphology of a healing bone-tendon interface: A histological observation in the sheep model. J Anat. 2007;210:318–327. 10. Galatz LM, Ball CM, Teefey SA, Middleton WD, Yamaguchi K. The outcome and repair integrity of completely arthroscopically repaired large and massive rotator cuff tears. J Bone Joint Surg Am. 2004;86:219–224.
88
11. Corry IS, Webb JM, Clingeleffer AJ, Pinczewski LA. Arthroscopic reconstruction of the anterior cruciate ligament: A comparison of patellar tendon autograft and four-strand hamstring tendon autograft. Am J Sports Med. 1999;27:444–454. 12. Tomita F, Yasuda K, Mikami S, Sakai T, Yamazaki S, Tohyama H. Comparisons of intraosseous graft healing between the doubled flexor tendon graft and the bone-patellar tendonbone graft in anterior cruciate ligament reconstruction. Arthroscopy 2001;17:461–476. 13. Leung KS, Qin L, Fu LK, Chan CW. A comparative study of bone to bone repair and bone to tendon healing in patellapatellar tendon complex in rabbits. Clin Biomech (Bristol, Avon) 2002;17:594–602. 14. Lui P, Zhang P, Chan K, Qin L. Biology and augmentation of tendon-bone insertion repair. J Orthop Surg Res. 2010;5:59. 15. Derwin KA, Baker AR, Spragg RK, Leigh DR, Iannotti JP. Commercial extracellular matrix scaffolds for rotator cuff tendon repair: Biomechanical, biochemical, and cellular properties. J Bone Joint Surg Am. 2006;88:2665–2672. 16. Barber FA, Herbert MA, Coons DA. Tendon augmentation grafts: Biomechanical failure loads and failure patterns. Arthroscopy 2006;22:534–538. 17. Walton JR, Bowman NK, Khatib Y, Linklater J, Murrell GA. Restore orthobiologic implant: Not recommended for augmentation of rotator cuff repairs. J Bone Joint Surg Am. 2007;89:786–791. 18. Valentin JE, Badylak JS, McCabe GP, Badylak SF. Extracellular matrix bioscaffolds for orthopaedic applications: A comparative histologic study. J Bone Joint Surg Am. 2006;88: 2673–2686. 19. Soler JA, Gidwani S, Curtis MJ. Early complications from the use of porcine dermal collagen implants (Permacol) as bridging constructs in the repair of massive rotator cuff tears: A report of 4 cases. Acta Orthop Belg. 2007;73:432–436. 20. Iannotti JP, Codsi MJ, Kwon YW, Derwin K, Ciccone J, Brems JJ. Porcine small intestine submucosa augmentation of surgical repair of chronic two-tendon rotator cuff tears: A randomized, controlled trial. J Bone Joint Surg Am. 2006;88:1238–1244. 21. Jungebluth P, Bader A, Baiguera S, et al. The concept of in vivo airway tissue engineering. Biomaterials 2012;33:4319–4326. 22. Chen F, Yoo JJ, Atala A. Acellular collagen matrix as a possible “off the shelf” biomaterial for urethral repair. Urology 1999;54:407–410. 23. Dahl SL, Koh J, Prabhakar V, Niklason LE. Decellularized native and engineered arterial scaffolds for transplantation. Cell Transplant. 2003;12:659–666. 24. Badylak SF. The extracellular matrix as a biologic scaffold material. Biomaterials 2007;28:3587–3593. 25. Badylak SF, Weiss DJ, Caplan A, Macchiarini P. Engineered whole organs and complex tissues. Lancet 2012;379:943–952. 26. Pridgen BC, Woon CY, Kim M, et al. Flexor tendon tissue engineering: Acellularization of human flexor tendons with preservation of biomechanical properties and biocompatibility. Tissue Eng Part C Methods 2011;17:819–828. 27. Woon CY, Pridgen BC, Kraus A, Bari S, Pham H, Chang J. Optimization of human tendon tissue engineering: Peracetic acid oxidation for enhanced reseeding of acellularized intrasynovial tendon. Plast Reconstr Surg. 2011;127:1107–1117. 28. Farnebo S, Woon CY, Bronstein JA, et al. Reconstruction of the tendon-bone insertion with decellularized tendon-bone composite grafts: comparison with conventional repair. J Hand Surg Am. E-published ahead of print November 15, 2013. 29. Bronstein JA, Woon CY, Farnebo S, et al. Tissue engineering of human tendon-bone interfaces: Physicochemical
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts decellularization of composite flexor tendon and bone. Plast Reconstr Surg. 2013;132:94–102. 30. Endress R, Woon CY, Farnebo SJ, et al. Tissue-engineered collateral ligament composite allografts for scapholunate ligament reconstruction: An experimental study. J Hand Surg Am. 2012;37:1529–1537. 31. Wolf MT, Daly KA, Brennan-Pierce EP, et al. A hydrogel derived from decellularized dermal extracellular matrix. Biomaterials 2012;33:7028–7038. 32. Whitlock PW, Seyler TM, Parks GD, et al. A novel process for optimizing musculoskeletal allograft tissue to improve safety, ultrastructural properties, and cell infiltration. J Bone Joint Surg Am. 2012;94:1458–1467. 33. Minami A, Ishii S, Ogino T, Oikawa T, Kobayashi H. Effect of the immunological antigenicity of the allogeneic tendons on tendon grafting. Hand 1982;14:111–119. 34. Getgood A, Bollen S. What tissue bankers should know about the use of allograft tendons and cartilage in orthopaedics. Cell Tissue Bank. 2010;11:87–97. 35. Raghavan SS, Woon CY, Kraus A, et al. Human flexor tendon tissue engineering: Decellularization of human flexor tendons reduces immunogenicity in vivo. Tissue Eng Part A 2012;18:796–805. 36. Baiguera S, Gonfiotti A, Jaus M, et al. Development of bioengineered human larynx. Biomaterials 2011;32:4433–4442. 37. Lomas RJ, Jennings LM, Fisher J, Kearney JN. Effects of a peracetic acid disinfection protocol on the biocompatibility and biomechanical properties of human patellar tendon allografts. Cell Tissue Bank. 2004;5:149–160. 38. Moffat KL, Sun WH, Pena PE, et al. Characterization of the structure-function relationship at the ligament-to-bone interface. Proc Natl Acad Sci USA 2008;105:7947–7952. 39. McAndrews PT, Arnoczky SP. Meniscal repair enhancement techniques. Clin Sports Med. 1996;15:499–510. 40. Thomopoulos S, Das R, Birman V, et al. Fibrocartilage tissue engineering: The role of the stress environment on cell morphology and matrix expression. Tissue Eng Part A 2011;17:1039–1053. 41. Nourissat G, Diop A, Maurel N, et al. Mesenchymal stem cell therapy regenerates the native bone-tendon junction
after surgical repair in a degenerative rat model. PLoS One 2010;5:e12248. 42. Ma CB, Kawamura S, Deng XH, et al. Bone morphoge netic proteins-signaling plays a role in tendon-to-bone healing: A study of rhBMP-2 and noggin. Am J Sports Med. 2007;35:597–604. 43. Thomopoulos S, Soslowsky LJ, Flanagan CL, et al. The effect of fibrin clot on healing rat supraspinatus tendon defects. J Shoulder Elbow Surg. 2002;11:239–247. 44. Ingram JH, Korossis S, Howling G, Fisher J, Ingham E. The use of ultrasonication to aid recellularization of acellular natural tissue scaffolds for use in anterior cruciate ligament reconstruction. Tissue Eng. 2007;13:1561–1572. 45. Jeremias Júnior SL, Camanho GL, Bassit AC, Forgas A, Ingham SJ, Abdalla RJ. Low-intensity pulsed ultrasound accelerates healing in rat calcaneus tendon injuries. J Orthop Sports Phys Ther. 2011;41:526–531. 46. Hays PL, Kawamura S, Deng XH, et al. The role of macrophages in early healing of a tendon graft in a bone tunnel. J Bone Joint Surg Am. 2008;90:565–579. 47. Killian ML, Cavinatto L, Galatz LM, Thomopoulos S. The role of mechanobiology in tendon healing. J Shoulder Elbow Surg. 2012;21:228–237. 48. Galatz LM, Charlton N, Das R, Kim HM, Havlioglu N, Thomopoulos S. Complete removal of load is detrimental to rotator cuff healing. J Shoulder Elbow Surg. 2009;18:669–675. 49. Hettrich CM, Rodeo SA, Hannafin JA, Ehteshami J, Shubin Stein BE. The effect of muscle paralysis using Botox on the healing of tendon to bone in a rat model. J Shoulder Elbow Surg. 2011;20:688–697. 50. Thomopoulos S, Williams GR, Soslowsky LJ. Tendon to bone healing: Differences in biomechanical, structural, and compositional properties due to a range of activity levels. J Biomech Eng. 2003;125:106–113. 51. Omae H, Zhao C, Sun YL, An KN, Amadio PC. Multilayer tendon slices seeded with bone marrow stromal cells: A novel composite for tendon engineering. J Orthop Res. 2009;27:937–942. 52. Cole DW, Ginn TA, Chen GJ, et al. Cost comparison of anterior cruciate ligament reconstruction: Autograft versus allograft. Arthroscopy 2005;21:786–790.
89
Background: Restoration of biomechanical strength following surgical reconstruction of tendon or ligament insertion tears is challenging because these injuries typically heal as fibrous scars. The authors hypothesize that injuries at the tendon-bone interface would benefit from reconstruction with decellularized composite tendon-bone grafts. Methods: Tendon-bone grafts were harvested from Sprague-Dawley rats. Grafts subjected to decellularization were compared histologically and biomechanically with untreated grafts ex vivo and in a new in vivo model. Wistar rats underwent Sprague-Dawley allograft reconstruction using a pair-matched design. The rats were killed at 2 or 4 weeks. B-cell and macrophage infiltration was determined using immunohistochemistry, and explants were tested biomechanically. Results: Decellularization resulted in a decrease in cells from 164 ± 61 (untreated graft) to 13 ± 7 cells per high-power field cells (p < 0.005) and a corresponding significant decrease in DNA content, and preserved scaffold architecture of the tendon-bone interface. Biomechanical comparison revealed no difference in failure load (p = 0.32), ultimate tensile stress (p = 0.76), or stiffness (p = 0.22) between decellularized grafts and untreated controls. Following in vivo reconstruction with tendon-bone interface grafts, decellularized grafts were stronger than untreated grafts at 2 weeks (p = 0.047) and at 4 weeks (p < 0.005). A persistent increase in B-cell and macrophage infiltration was observed in both the capsule surrounding the tendon-bone interface and the tendon substance in untreated controls. Conclusion: Decellularized tendon-bone grafts display better biomechanical properties at early healing time points and a decreased immune response compared with untreated grafts in vivo. (Plast. Reconstr. Surg. 133: 79, 2014.)
I
njuries to the site where tendons and ligaments insert onto bone are common and constitute a significant reconstructive challenge. Common examples include rupture of flexor tendon insertion at the distal phalanx,1 rupture of tendons in the rotator cuff,2,3 and rupture of the anterior cruciate ligament.4 Current reconstructive strategies involve hardware, suture anchors, and drill holes to approximate tendons and ligaments to From the Division of Plastic Surgery, Veterans Affairs Palo Alto Health Care System, and the Division of Plastic Surgery, Stanford University Medical Center. Received for publication April 17, 2012; accepted June 5, 2013. Presented at the 67th Annual Meeting of the American Society for Surgery of the Hand, in Chicago, Illinois, September 6 through 8, 2012. Copyright © 2013 by the American Society of Plastic Surgeons DOI: 10.1097/01.prs.0000436823.64827.a0
bone. However, the native tendon-bone interface is typically never fully reconstituted after surgery. Although gross healing may be present,5 microscopic inspection of the reconstructed tendonbone interface will reveal fibrous scar tissue with inferior biomechanical properties compared with the native insertion.6–9 Increased risk for failure ensues and, as a result, current surgical repair techniques have high reinjury rates. Up to 94 percent Disclosure: None of the authors has a financial interest in any of the products, devices, or drugs mentioned in this article.
This work was supported by the plastic surgery foundation.
www.PRSJournal.com
79
Plastic and Reconstructive Surgery • January 2014 of repaired massive rotator cuff tears fail,10 and 56 percent of anterior cruciate ligament reconstruction patients experience knee pain 1 year after surgery.11 Conceptually, the use of a decellularized tendon-bone extracellular matrix construct to replace a tendon-bone interface deficit is an attempt to replace the weaker tendon-to-bone healing with bone-to-bone and tendon-to-tendon healing, which is regarded as stronger because it involves the approximation of tissues with equivalent elastic moduli.12–14 Currently available extracellular matrix scaffolds approved for augmentation of tendon healing are primarily of animal origin.15,16 As with all nonhuman tissue grafts, concerns regarding the degree of decellularization and strong immune responses toward these xenogeneic materials have been raised.17–20 Human cadaveric allograft tissue represents an underused, widely available tissue source that may provide an alternative scaffold option for tissue-engineered composite grafts.21–25 Previous research on decellularized human flexor tendons26,27 and multiple composite tissue scaffolds (tendon-bone interface and ligament-bone grafts)28–30 may provide alternative treatment options in reconstructive surgery. These allografts contain extracellular matrix proteins such as collagen, elastin, and proteoglycans that are crucial for cell attachment, migration, and proliferation.31,32 It is believed that the decellularized extracellular matrix is insoluble and conserved across species and therefore does not trigger an immune response on reimplantation, compared with cellular debris, which is much more immunogenic.33,34 Therefore, it is important that allograft scaffolds used for upper extremity reconstruction be decellularized.24–26,35 A previous model of human flexor tendons has also shown that chemical decellularization reduces cellular material to a minimum (94 percent decrease of cellular content)26 and that this process eliminates the major histocompatibility complexes that are responsible for immune recognition33,34,36 and thereby reduces the immune response in vivo.29,35 The rat Achilles tendon-bone interface model was used to represent the flexor digitorum
profundus insertion at the distal phalanx. The rat Achilles tendon was chosen because (1) its superficial location enables easy access to surgical transection and subsequent reconstruction, (2) the rodent calcaneus is a bone that is large enough to tolerate clinical hardware (microdrill and screw), (3) it allows for comparison in pair-matched analysis with the contralateral side, and (4) it is of a similar order of magnitude (similar size ratio) as human digital flexors. In this article, we aim to show that decellularization, a necessary step for addressing immunogenicity and disease transmission, does not adversely affect scaffold integrity and strength properties after treatment. We hypothesize that (1) composite grafts can successfully be decellularized with preservation of biomechanical properties and structure, (2) decellularization substantially decreases the inflammatory response after in vivo allograft reconstruction, and (3) decellularized grafts are stronger at the early healing time points than untreated grafts in vivo.
MATERIALS AND METHODS Animals and Experimental Groups Eighty-six Sprague-Dawley (mean weight, 280 g; Charles River Laboratories, Willimantic, Conn.) Achilles tendon–calcaneus composite tendon-bone interface grafts were harvested. The grafts were divided into four groups and assessed with histology and biomechanical strength testing (Table 1). Grafts used as untreated allograft controls were snap-frozen in saline to −70°C after harvest and remained frozen until further use. Decellularization Targeted ultrasonication (VC505; Sonics & Materials, Newtown, Conn.) against the tendonbone interface of the grafts was performed in a chilled water bath (10 minutes; total, 64,800 J).29 Buffered 5% peracetic acid (Sigma, St. Louis, Mo.), 1% ethylenediaminetetraacetic acid, and 2% sodium dodecyl sulfate in 1% ethylenediaminetetraacetic acid was used for chemical
Table 1. Data for the Four Groups Group In vitro 1 2 In vivo 3 4
Description
Total No.
Histology
Strength
DNA Content
Untreated controls, SD grafts Decellularized grafts, SD grafts
19 23
5 9
9 9
5 5
Untreated controls, SD grafts in Wistar rats Decellularized grafts, SD grafts in Wistar rats
22 22
2 (2 wk) + 2 (4 wk) 2 (2 wk) + 2 (4 wk)
9 (2 wk) + 9 (4 wk) 9 (2 wk) + 9 (4 wk)
0 0
SD, Sprague-Dawley.
80
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts decellularization according to a modification of a previously established protocol.26 All grafts were rinsed with deionized water to ensure that grafts were chemical free before implantation. Our decellularization protocol was used as the sole sterilizing agent, as previous studies have shown that 5% peracetic acid and detergents efficiently minimize disease transmission.32,37 In Vitro Experiments Histologic Analysis Untreated and decellularized tendon-bone interface grafts were fixed in 10% formalin and decalcified overnight (Decal; Decal Chemical Corp., Tallman, N.Y.). Specimens were embedded in paraffin; sectioned (6 μm); and stained with hematoxylin and eosin, SYTOGreen fluorescent nucleic acid stain (Invitrogen, Carlsbad, Calif.), and Masson’s trichrome to determine the adequacy of decellularization and preservation of collagenous architecture. Cell Counts Using hematoxylin and eosin–stained sections, cell counts were performed by two independent observers. Eight random representative high-power field (20×) (Nikon TS100; Nikon, Melville, N.Y.) images were centered on the tendon-bone interface with a superscript grid, to ensure that the evaluated area examined remained the same in all samples. DNA Content Untreated and decellularized tendon-bone interface grafts were trimmed, lyophilized, and weighed. Total DNA was then isolated from each sample
(DNAeasy; Qiagen, Hilden, Germany). DNA content was calculated from the absorption at 280 nm analyzed in triplicate using a Spectrophotometer (ND-1000; NanoDrop, Wilmington, Del.) and normalized to the initial dry weight of the sample. Biomechanical Testing A pair-matched design was used to compare material properties between in vitro groups and between in vivo groups (untreated versus decellularized). Harvested grafts were secured in a custom-made clamp (Fig. 1) with the dissected foot locked securely under a metal bar and the tendon end pulled to failure (ultimate failure load) of the graft (MTS 858; MTS Systems Corp., Eden Prairie, Minn.). Orthogonal photographs together with a known scale (digital caliper) were obtained using a digital camera (resolution, 2592 × 1936 pixels) at a fixed distance to measure two dimensions of the tendon. The cross-sectional area was calculated using an elliptical assumption (ImageJ software; National Institutes of Health, Bethesda, Md.) and used to calculate the ultimate tensile stress (force divided by cross-sectional area). The constructs were pulled until failure at a rate of 0.5 mm/minute. The MTS system recorded force and displacement data continuously during each test. Stiffness was calculated using the gradient of the force-displacement curve during the linear phase of deformation. In Vivo Experiments Surgical Technique We used a rat allograft model (implantation of Sprague-Dawley allografts into Wistar rats; mean
Fig. 1. Biomechanical setup. (Left) Degloved rat foot with intramuscular portion of the Achilles tendon (1), Achilles tendon (2), and calcaneus bone (3). (Center and right) Photographs of rat foot in custom-made clamp (a) with the tibia locked securely outside the holding block (b) in two positions perpendicular to each other. The tendon ends were fastened between sheets of sandpaper using commercial cyanoacrylate glue and clamped between the crossheads of a servo-controlled hydraulic materials testing system (MTS 858). Arrows show direction of pull.
81
Plastic and Reconstructive Surgery • January 2014 weight, 320 g) to represent the scenario of clinical allograft reconstruction of the tendon-bone interface. The procedure is described in the legend for Figure 2. After surgery, all animals were allowed immediate mobilization without the need for a cast. All animal procedures were performed with an institutional animal care and use committee– approved animal protocol. Immunohistochemistry Tendon-bone interface grafts were harvested at 2 and 4 weeks for histologic analysis. For B-cell staining, a goat anti-rat CD20 primary antibody was used (Santa Cruz Biotechnology, Santa Cruz, Calif.). For macrophage staining, a mouse antirat CD68 primary antibody was used (Santa Cruz Biotechnology). Rat spleen was used as a positive control. Micro–Computed Tomography High-resolution (10.5-μm voxel size) postoperative imaging with micro–computed tomography (Viva CT 75; Scanco Medical AG, Brüttisellen, Switzerland) was performed for closer visualization of screw placement in the graft and calcaneus bone in four of the animals (Fig. 2). Statistical Analysis Data were reported as mean ± SD. Comparisons of continuous data across groups were performed
using paired t test. The significance level was set at p < 0.05. All analyses were performed using PASW Statistics for Windows, Version 18.0 (SPSS, Inc., Chicago, Ill.).
RESULTS Histology (In Vitro) Untreated controls demonstrated cellularity in tendon, fibrocartilage, and bone (Fig. 3, above, left, and center, left). Following decellularization, there was reduced cellularity in all parts of the constructs, including the fibrocartilage zone of the tendon-bone interface, as shown by hematoxylin and eosin– and Masson trichrome– stained images (Fig. 3, above, right, and center, right)) and SYTOGreen nucleic acid stain (Fig. 3, below, right). There were no visible changes in the morphology of the tendon, fibrocartilage, or bone following treatment, indicating microstructural preservation with almost complete removal of visible cells. Cell Counts (In Vitro) Physicochemical decellularization resulted in significant reduction of visible cells as shown with cell counts. Average cell counts decreased from 164 ± 61 cells/high-power field (untreated grafts)
Fig. 2. The Achilles tendon was sharply released from the calcaneus bone, and the posterior tuberosity of the calcaneus was dissected free. The bone was cut with an oscillating microsagittal saw approximately 3 mm distal to the tendon insertion (Stryker, Kalamazoo, Mich.) and thread holes were drilled (0.75 mm; Medartis AG, Basel, Switzerland) in the center of the bone. The graft bone segments were then drilled (0.75 mm; Medartis) and fixed with 0.9 mm (diameter) × 10 mm (length) self-tapping microscrews (Medartis) (left). The proximal tendon was cut proximally at the musculotendinous junction and sutured to the free end of the transected Achilles tendon using a 4-0 Prolene (Ethicon, Inc., Somerville, N.J.) locking stitch. Tendon-bone interface composite construct with 0.9-mm microscrew in place (left). Composite construct implanted in rat as an allograft (center). Micro–computed tomographic image of graft and screw in calcaneus bone (right).
82
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 3. Physicochemical decellularization resulted in significant reduction of visible cells and nucleic material, with microstructural preservation. Hematoxylin and eosin–stained untreated tendon-bone interface (above, left) and decellularized tendon-bone interface (above, right) (original magnification, × 10 for both). Masson trichrome–stained untreated tendon-bone interface (center, left) and decellularized tendon-bone interface (center, right) (original magnification, × 20 for both). Hematoxylin and eosin–stained decellularized tendon-bone interface (below, left) (original magnification, × 20) and SYTOgreen nuclear stain decellularized tendon-bone interface (below, right) (original magnification, × 20). T, tendon; FC, fibrocartilage; B, bone.
to 13 ± 7 cells/high-power field (decellularized grafts) (p = 0.0004). DNA Content The mean DNA content of the untreated grafts was 57.9 ± 7.5 ng/mg DNA/dry weight, compared with 21.7 ± 9.0 ng/mg DNA/dry weight, a significant decrease of 63 percent (p < 0.005). Biomechanical Testing (In Vitro and In Vivo) All biomechanical data are listed in Table 2, including mode of failure.
Animal Behavior (In Vivo) In the first week, the reconstructed animals showed a slight bilateral swelling around their feet. However, they started immediately to bear weight on both legs and walked around their cages with only a slight limp that disappeared after the first 2 weeks. Histology (In Vivo) At 2 weeks after implantation, increases in cellular infiltration and vascularization of the capsule were seen in both groups. A clear
83
Plastic and Reconstructive Surgery • January 2014 Table 2. Comparison of Data* Group In vitro 1 2 In vivo 3 4 3 4
Ultimate Failure Load (N)
Ultimate Tensile Stress (MPa)
Stiffness (N/mm)
Untreated controls Decellularized grafts
75.7 ± 18.5 69.5 ± 6.0 (NS)
10.2 ± 4.9 10.7 ± 3.0 (NS)
17.9 ± 5.5 15.2 ± 3.0 (NS)
2 wk, untreated grafts 2 wk, decellularized grafts 4 wk, untreated grafts 4 wk, decellularized grafts
23.3 ± 9.1 31.7 ± 7.4† 22.7 ± 9.6 46.9 ± 12.7‡
Description
1.2 ± 0.6 1.9 ± 0.6† 1.1 ± 0.5 2.3 ± 0.7‡
6.4 ± 2.0 7.1 ± 2.7 (NS) 4.3 ± 1.1 7.6 ± 0.9‡
Mode of Failure All tendon All tendon TBI, 5; tendon, 3; bone, 1 TBI, 4; tendon, 4; bone, 1 TBI, 6; tendon, 2; bone, 1 TBI, 7; tendon, 1; bone, 1
NS, nonsignificant (p > 0.05); TBI, tendon-bone interface. *Pair-matched samples: untreated grafts versus decellularized grafts in vitro, at 2 weeks, and at 4 weeks. In all cases of bone failure [n = 4 (11%)], a fracture of the drill canal through the graft bone was observed. An asymmetrical drilling of the graft canal was seen in three of these cases as a possible explanation; no slipping of the screws was seen in any of the grafts. †p < 0.05. ‡p < 0.005.
noninflammatory cellular infiltration of the tendon was seen only in decellularized grafts. Immunohistochemistry showed that macrophages were confined to the capsule and absent in the tendon substance, indicating the classic and sequential pattern of inflammatory cell accumulation as seen postoperatively after reconstruction with an allograft. Immunohistochemistry of B cells, in contrast, displayed a clear difference between the groups, with infiltration throughout both the capsule and the tendon substance in the untreated grafts (Figs. 4 and 5). This B-cell infiltration of the tendon, and the capsule, persisted and increased after 4 weeks, compared with the decellularized grafts. Also, macrophage infiltration of the capsule remained more prominent in the untreated grafts compared with the decellularized grafts (Fig. 6). The inflammatory cells appeared to invade the tendons in “sheets” physically separating the collagen fibers. Over time, more spindleshaped cells appeared in both groups, indicating a normal repopulation of the tissues.
DISCUSSION This article highlights the need for allograft decellularization to decrease graft immunogenicity and to accelerate postoperative healing in reconstructive surgery. Also, it demonstrates how decellularized composite tissue grafts offer an alternative reconstructive solution to complex tendon-bone ruptures in vivo. The use of decellularized tendon-bone constructs may thus challenge the existing paradigm of suture-based tendon-bone interface repair, suggesting that the damaged tendon-bone interface can be replaced with a biocompatible decellularized “off-the-shelf” tendon-bone interface construct. The four-zone tendon-bone interface transmits loads and decreases the concentration of
84
stress at the attachment zone.38 Collagen II–producing chondrocytes of the fibrocartilage are likely to be a key element in this biomechanical function38; however, they typically do not regenerate after injury, which precludes regeneration of the fibrocartilage zone.5,39 Therefore, previous studies have aimed at improving the healing of the fibrocartilage by inducing stress,40 local application of stem cells,41 use of growth factors,42 and use of fibrin glue.43 We have previously established a protocol to decellularize single tissue flexor tendon scaffolds using a combination of detergents26 and increasing microstructural scaffold porosity with peracetic acid.27 However, the fibrocartilage zone is only partially decellularized using these protocols.36 This is because the chondrocytes within the fibrocartilage are deeply embedded within their lacunae in a tight-knit, partially mineralized, stiff tendon fiber matrix. We have therefore added a mechanical component to the protocol to increase detergent penetration in the hard-to-penetrate fibrocartilage zone.29 Targeted ultrasonication as a first step is effective because it increases tissue porosity and mechanically disrupts cell membranes before extraction of nuclear material by the detergent solution.44 Microscopic analysis revealed a nearly complete reduction (92 percent) of cellular elements and nuclear material removed during the process of decellularization without collateral damage to the extracellular matrix structure. It has recently been shown in a flexor tendon xenograft model that decellularization is crucial to limiting an immune response in vivo and that a vigorous immune reaction can rapidly compromise biomechanical strength.35 An absent or much-attenuated immune response will be beneficial during initiation of early mobilization. The biomechanical
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 4. Hematoxylin and eosin staining (original magnification, × 20) showing tendon substance of untreated graft (above, left) and decellularized graft (above, right) at 2 weeks, and untreated graft (second row, left) and decellularized graft (second row, right) at 4 weeks. Although it is impossible to distinguish donor cell persistence in the untreated grafts (above, left and second row, left) from host cell infiltration, the greater number of cells in the untreated grafts is likely to be of inflammatory origin, based on their abundance seen only in the untreated grafts and also based on their morphology. Remaining tenocytes typically present as elongated cells (arrowheads, below, right) whereas the macrophages and lymphocytes are round (arrowheads, below, right). High-power (original magnification, × 40) insets of untreated graft at 2 weeks (third row, left) with
85
Plastic and Reconstructive Surgery • January 2014
Fig. 5. Immunohistochemistry staining (original magnification, × 20) of CD20+ lymphocytes in the tendinous part of an untreated graft (above, left) and decellularized graft (above, right) at 2 weeks, untreated graft (below, left), and decellularized graft (below, right) at 4 weeks.
strength of our decellularized construct is comparable to what was observed in previous studies on the Achilles tendon insertion in the calcaneus bone.41,45 Also, and more importantly, the ultimate failure load is comparable to the strength of a pullout suture model where healing was augmented by mesenchymal stem cells.41 However, future studies, using this model will be needed to determine its benefits compared with a conventional suture pullout repair and when this extracellular matrix scaffold is reseeded with pluripotent cells. The Sprague-Dawley–to-Wistar rat allograft model was intended to simulate clinical allogenic composite tissue transplantation. Because of species inbreeding, single-species rat isograft reconstruction resembles autograft replacement in design and may not sufficiently replicate the immune response expected in clinical allograft transplantation between patients. Reliable markers to evaluate the degree of immune response were chosen to Fig. 4. (Continued) polymorphonuclear infiltration in “sheaths” separating the tendon weave (asterisks), and surrounding vessels (third row, right). Untreated donor tendon (dT) at the region of tendon suture of the recipient rat Achilles tendon (rT) at 20× (below, left) and 40× (below, right) magnification illustrating the vast inflammatory response toward untreated graft tendon but not recipient tendon.
86
allow for adequate explanation of biomechanical testing results. CD20 and CD68 cell surface markers were used to identify B-cell and macrophage invasion at time points chosen to reflect the early inflammatory phase of postoperative healing. Our results concur with previous studies that have shown that macrophages are present during the early inflammatory phase of tendon and tendon-bone healing. These macrophages are known to produce proinflammatory mediators that may have a negative effect on tissue regeneration and strength, especially in the early “destructive” phase of inflammatory healing. Classically activated macrophages release proinflammatory cytokines such as interleukin-1, interluekin-6, and tumor necrosis factor-α, and matrix proteases, which aid in cell chemotaxis through the inflammatory milieu. It has been shown that tendon-bone healing is enhanced through reestablishment of collagen fiber continuity, new bone formation, and the overall degree of interface remodeling if macrophages are depleted from the injured zone.46 B-cell infiltration of the tendon core and capsule indicates that the humoral pathway of the immune system is also activated, likely leading to collagen degradation and compromised tendon structure. The status of grafts in this study resembles what we previously described
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts
Fig. 6. Immunohistochemistry (original magnification, × 20) staining of CD20+ lymphocytes at 4 weeks, showing an increased number of inflammatory cells in the capsule surrounding the tendon-bone interface in the untreated group (above, left) compared with the decellularized group (above, right). CD68+ macrophages at 4 weeks, showing an increased number of inflammatory cells in the capsule (c) and tendon (t) surrounding the tendon-bone interface in the untreated group (below, left) compared with the decellularized group (below, right).
in decellularized human flexor tendons used as xenografts in subcutaneous pockets of rats, where remaining major histocompatibility complexes in untreated tendons triggered a similar strong inflammatory response.35 Compared with untreated grafts, decellularized grafts were 1.4 times stronger at 2 weeks, and 2.0 times stronger at 4 weeks. These results can be explained by the weaker inflammatory response to decellularized grafts compared with untreated grafts, leading to superior biomechanical properties in the early healing period. Early strength will allow for initiation of early active motion protocols, facilitating earlier recovery of function and return to activities, and also impedes adhesion formation.40,47 A fine balance must be achieved between unloading and overloading of the tendon-bone interface. Earlier studies of conventional reattachment of the tendon to bone indicate that early mobilization is beneficial for extracellular matrix production at the repair site.48,49 However, excessive motion and load can also be detrimental to healing, as this extracellular matrix is composed of inferior scar-like material.50 Conservative postoperative protocols involving lengthy immobilization
are therefore often advocated. These protocols lead to gradual accumulation of tendon tensile strength at the expense of gliding. A closer analysis of pair-matched constructs revealed a significantly higher stiffness in the decellularized grafts at 4 weeks in vivo, compared with the untreated grafts. The stiffness of the untreated grafts continues to drop over time (from 6.4 N/mm to 4.3 N/mm) without a corresponding drop in ultimate failure load or ultimate tensile strength, whereas stiffness in the decellularized grafts improves only slightly (from 7.1 N/ mm to 7.6 N/mm) despite a 48 percent increase in ultimate failure load (from 31.7 N/mm to 46.9 N). The clinical impact of these discrepancies remains unknown, and we hesitate to draw conclusions for two reasons. First, we see a potential risk that when grafts are ruptured at various locations (e.g., tendon left and tendon-bone insertion right), the stress or stiffness would not be comparable. Second, stress values from composite tissue testing must be regarded with caution when composite tissue grafts are used, as the cross-sectional area at the side of insertion is very irregular, variable, and not necessarily elliptical.
87
Plastic and Reconstructive Surgery • January 2014 CONCLUSIONS Decellularized tendon-bone interface grafts possess ideal characteristics of an allograft scaffold—basic tendon and tendon-bone interface structure, native extracellular matrix architecture, and optimal biocompatibility.51 In the future, these grafts may be used to precisely match deficient tendon and complex tendon-bone insertion tears without donor-site morbidity.52 James Chang, M.D. Division of Plastic Surgery Stanford University Medical Center 770 Welch Road, Suite 400 Stanford, Calif. 94305 [email protected]
ACKNOWLEDGMENT
This project was funded by a Plastic Surgery Foundation grant, an American Foundation for Surgery of the Hand/American Society for Surgery of the Hand grant, and a Veterans Affairs Rehabilitation Research and Development Merit Review grant. REFERENCES 1. Moiemen NS, Elliot D. Primary flexor tendon repair in zone 1. J Hand Surg Br. 2000;25:78–84. 2. Lehman C, Cuomo F, Kummer FJ, Zuckerman JD. The incidence of full thickness rotator cuff tears in a large cadaveric population. Bull Hosp Jt Dis. 1995;54:30–31. 3. Yamaguchi K, Ditsios K, Middleton WD, Hildebolt CF, Galatz LM, Teefey SA. The demographic and morphological features of rotator cuff disease: A comparison of asymptomatic and symptomatic shoulders. J Bone Joint Surg Am. 2006;88:1699–1704. 4. Brown CH Jr, Carson EW. Revision anterior cruciate ligament surgery. Clin Sports Med. 1999;18:109–171. 5. Thomopoulos S, Genin GM, Galatz LM. The development and morphogenesis of the tendon-to-bone insertion: What development can teach us about healing. J Musculoskelet Neuronal Interact. 2010;10:35–45. 6. Liu CF, Aschbacher-Smith L, Barthelery NJ, Dyment N, Butler D, Wylie C. What we should know before using tissue engineering techniques to repair injured tendons: A developmental biology perspective. Tissue Eng Part B Rev. 2011;17:165–176. 7. Forward AD, Cowan RJ. Experimental suture of tendon to bone. Surg Forum 1960;11:458–460. 8. Galatz LM, Rothermich SY, Zaegel M, Silva MJ, Havlioglu N, Thomopoulos S. Delayed repair of tendon to bone injuries leads to decreased biomechanical properties and bone loss. J Orthop Res. 2005;23:1441–1447. 9. Newsham-West R, Nicholson H, Walton M, Milburn P. Long-term morphology of a healing bone-tendon interface: A histological observation in the sheep model. J Anat. 2007;210:318–327. 10. Galatz LM, Ball CM, Teefey SA, Middleton WD, Yamaguchi K. The outcome and repair integrity of completely arthroscopically repaired large and massive rotator cuff tears. J Bone Joint Surg Am. 2004;86:219–224.
88
11. Corry IS, Webb JM, Clingeleffer AJ, Pinczewski LA. Arthroscopic reconstruction of the anterior cruciate ligament: A comparison of patellar tendon autograft and four-strand hamstring tendon autograft. Am J Sports Med. 1999;27:444–454. 12. Tomita F, Yasuda K, Mikami S, Sakai T, Yamazaki S, Tohyama H. Comparisons of intraosseous graft healing between the doubled flexor tendon graft and the bone-patellar tendonbone graft in anterior cruciate ligament reconstruction. Arthroscopy 2001;17:461–476. 13. Leung KS, Qin L, Fu LK, Chan CW. A comparative study of bone to bone repair and bone to tendon healing in patellapatellar tendon complex in rabbits. Clin Biomech (Bristol, Avon) 2002;17:594–602. 14. Lui P, Zhang P, Chan K, Qin L. Biology and augmentation of tendon-bone insertion repair. J Orthop Surg Res. 2010;5:59. 15. Derwin KA, Baker AR, Spragg RK, Leigh DR, Iannotti JP. Commercial extracellular matrix scaffolds for rotator cuff tendon repair: Biomechanical, biochemical, and cellular properties. J Bone Joint Surg Am. 2006;88:2665–2672. 16. Barber FA, Herbert MA, Coons DA. Tendon augmentation grafts: Biomechanical failure loads and failure patterns. Arthroscopy 2006;22:534–538. 17. Walton JR, Bowman NK, Khatib Y, Linklater J, Murrell GA. Restore orthobiologic implant: Not recommended for augmentation of rotator cuff repairs. J Bone Joint Surg Am. 2007;89:786–791. 18. Valentin JE, Badylak JS, McCabe GP, Badylak SF. Extracellular matrix bioscaffolds for orthopaedic applications: A comparative histologic study. J Bone Joint Surg Am. 2006;88: 2673–2686. 19. Soler JA, Gidwani S, Curtis MJ. Early complications from the use of porcine dermal collagen implants (Permacol) as bridging constructs in the repair of massive rotator cuff tears: A report of 4 cases. Acta Orthop Belg. 2007;73:432–436. 20. Iannotti JP, Codsi MJ, Kwon YW, Derwin K, Ciccone J, Brems JJ. Porcine small intestine submucosa augmentation of surgical repair of chronic two-tendon rotator cuff tears: A randomized, controlled trial. J Bone Joint Surg Am. 2006;88:1238–1244. 21. Jungebluth P, Bader A, Baiguera S, et al. The concept of in vivo airway tissue engineering. Biomaterials 2012;33:4319–4326. 22. Chen F, Yoo JJ, Atala A. Acellular collagen matrix as a possible “off the shelf” biomaterial for urethral repair. Urology 1999;54:407–410. 23. Dahl SL, Koh J, Prabhakar V, Niklason LE. Decellularized native and engineered arterial scaffolds for transplantation. Cell Transplant. 2003;12:659–666. 24. Badylak SF. The extracellular matrix as a biologic scaffold material. Biomaterials 2007;28:3587–3593. 25. Badylak SF, Weiss DJ, Caplan A, Macchiarini P. Engineered whole organs and complex tissues. Lancet 2012;379:943–952. 26. Pridgen BC, Woon CY, Kim M, et al. Flexor tendon tissue engineering: Acellularization of human flexor tendons with preservation of biomechanical properties and biocompatibility. Tissue Eng Part C Methods 2011;17:819–828. 27. Woon CY, Pridgen BC, Kraus A, Bari S, Pham H, Chang J. Optimization of human tendon tissue engineering: Peracetic acid oxidation for enhanced reseeding of acellularized intrasynovial tendon. Plast Reconstr Surg. 2011;127:1107–1117. 28. Farnebo S, Woon CY, Bronstein JA, et al. Reconstruction of the tendon-bone insertion with decellularized tendon-bone composite grafts: comparison with conventional repair. J Hand Surg Am. E-published ahead of print November 15, 2013. 29. Bronstein JA, Woon CY, Farnebo S, et al. Tissue engineering of human tendon-bone interfaces: Physicochemical
Volume 133, Number 1 • Tissue Engineering of Tendon-Bone Grafts decellularization of composite flexor tendon and bone. Plast Reconstr Surg. 2013;132:94–102. 30. Endress R, Woon CY, Farnebo SJ, et al. Tissue-engineered collateral ligament composite allografts for scapholunate ligament reconstruction: An experimental study. J Hand Surg Am. 2012;37:1529–1537. 31. Wolf MT, Daly KA, Brennan-Pierce EP, et al. A hydrogel derived from decellularized dermal extracellular matrix. Biomaterials 2012;33:7028–7038. 32. Whitlock PW, Seyler TM, Parks GD, et al. A novel process for optimizing musculoskeletal allograft tissue to improve safety, ultrastructural properties, and cell infiltration. J Bone Joint Surg Am. 2012;94:1458–1467. 33. Minami A, Ishii S, Ogino T, Oikawa T, Kobayashi H. Effect of the immunological antigenicity of the allogeneic tendons on tendon grafting. Hand 1982;14:111–119. 34. Getgood A, Bollen S. What tissue bankers should know about the use of allograft tendons and cartilage in orthopaedics. Cell Tissue Bank. 2010;11:87–97. 35. Raghavan SS, Woon CY, Kraus A, et al. Human flexor tendon tissue engineering: Decellularization of human flexor tendons reduces immunogenicity in vivo. Tissue Eng Part A 2012;18:796–805. 36. Baiguera S, Gonfiotti A, Jaus M, et al. Development of bioengineered human larynx. Biomaterials 2011;32:4433–4442. 37. Lomas RJ, Jennings LM, Fisher J, Kearney JN. Effects of a peracetic acid disinfection protocol on the biocompatibility and biomechanical properties of human patellar tendon allografts. Cell Tissue Bank. 2004;5:149–160. 38. Moffat KL, Sun WH, Pena PE, et al. Characterization of the structure-function relationship at the ligament-to-bone interface. Proc Natl Acad Sci USA 2008;105:7947–7952. 39. McAndrews PT, Arnoczky SP. Meniscal repair enhancement techniques. Clin Sports Med. 1996;15:499–510. 40. Thomopoulos S, Das R, Birman V, et al. Fibrocartilage tissue engineering: The role of the stress environment on cell morphology and matrix expression. Tissue Eng Part A 2011;17:1039–1053. 41. Nourissat G, Diop A, Maurel N, et al. Mesenchymal stem cell therapy regenerates the native bone-tendon junction
after surgical repair in a degenerative rat model. PLoS One 2010;5:e12248. 42. Ma CB, Kawamura S, Deng XH, et al. Bone morphoge netic proteins-signaling plays a role in tendon-to-bone healing: A study of rhBMP-2 and noggin. Am J Sports Med. 2007;35:597–604. 43. Thomopoulos S, Soslowsky LJ, Flanagan CL, et al. The effect of fibrin clot on healing rat supraspinatus tendon defects. J Shoulder Elbow Surg. 2002;11:239–247. 44. Ingram JH, Korossis S, Howling G, Fisher J, Ingham E. The use of ultrasonication to aid recellularization of acellular natural tissue scaffolds for use in anterior cruciate ligament reconstruction. Tissue Eng. 2007;13:1561–1572. 45. Jeremias Júnior SL, Camanho GL, Bassit AC, Forgas A, Ingham SJ, Abdalla RJ. Low-intensity pulsed ultrasound accelerates healing in rat calcaneus tendon injuries. J Orthop Sports Phys Ther. 2011;41:526–531. 46. Hays PL, Kawamura S, Deng XH, et al. The role of macrophages in early healing of a tendon graft in a bone tunnel. J Bone Joint Surg Am. 2008;90:565–579. 47. Killian ML, Cavinatto L, Galatz LM, Thomopoulos S. The role of mechanobiology in tendon healing. J Shoulder Elbow Surg. 2012;21:228–237. 48. Galatz LM, Charlton N, Das R, Kim HM, Havlioglu N, Thomopoulos S. Complete removal of load is detrimental to rotator cuff healing. J Shoulder Elbow Surg. 2009;18:669–675. 49. Hettrich CM, Rodeo SA, Hannafin JA, Ehteshami J, Shubin Stein BE. The effect of muscle paralysis using Botox on the healing of tendon to bone in a rat model. J Shoulder Elbow Surg. 2011;20:688–697. 50. Thomopoulos S, Williams GR, Soslowsky LJ. Tendon to bone healing: Differences in biomechanical, structural, and compositional properties due to a range of activity levels. J Biomech Eng. 2003;125:106–113. 51. Omae H, Zhao C, Sun YL, An KN, Amadio PC. Multilayer tendon slices seeded with bone marrow stromal cells: A novel composite for tendon engineering. J Orthop Res. 2009;27:937–942. 52. Cole DW, Ginn TA, Chen GJ, et al. Cost comparison of anterior cruciate ligament reconstruction: Autograft versus allograft. Arthroscopy 2005;21:786–790.
89
