MOLECULAR ENHANCEMENT OF EXTRACELLULAR MATRIX AND METHODS

OF USE

CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims the benefit of U.S. Provisional

Application No. 61/065,527 filed February 13, 2008, the entire disclosure of which is incorporated herein by reference.

BACKGROUND OF THE INVENTION

[0002] The rotator cuff comprises four muscles that control shoulder movement and their associated tendons by which the muscles are attached to the humeral head (the end of the humerus that forms part of the shoulder joint). Together with the joint capsule, these muscles and tendons form a cuff that surrounds the head of the humerus. The tendons are thick, ribbon-like structures having high tensile strength. They not only hold the shoulder muscles and the humerus together, but also transmit forces exerted by those muscles to induce corresponding movement of the humerus. In a healthy subject, the tendons are fully attached at one end to, and envelop, a portion of the humeral-head surface. When a rotator cuff tear occurs, the joint capsule and one or more of the tendons become(s) partially or entirely torn away from the humeral head, creating both a painful and a functionally debilitating condition.

[0003] Current treatment for rotator-cuff tears is to suture the torn tendon back to the bone of the humeral head. The sutures hold the tendon in contact with the bone, preferably long enough for the tendon to heal to the bone and form a bridge that will re-establish the tendon-bone connection and restore normal function. The sutures that are used possess sufficient tensile strength to retain the tendon and bone together during the healing process. However, the tendon is a fibrous tissue that can be torn by the sutures. Commonly, the sutures will align with the fascicular structure of the tendon and tear right through it under sufficient tensile force, thus undoing the surgical repair before tendon-to-bone healing is complete. The sutures can also tear through the bone under sufficient force, particularly in older subjects who form the bulk of rotator-cuff-tear patients and whose bones tend to be more osteoporotic.

[0004] In fact, rotator-cuff tears affect 40% or more of those over age

60 and cost the US economy approximately $3 billion per year. The repair failure rate of large to massive rotator cuff tears ranges from 20 to 90%. High re-tear rates are a result of mechanical factors (e.g., tear size, repair technique, rehabilitation protocol) as well as biologic factors (e.g., age, tear chronicity, tendon quality, disease) that may compromise the patients' intrinsic capacity to heal. All of these factors may also contribute to the propensity of the sutures to tear through the tendon and bone before healing is complete, thus contributing to the re-tear rate. Hence, there is a need for repair strategies that provide adequate strength as well as stimulate and enhance healing potential.

BRIEF SUMMARY OF THE INVENTION

[0005] A composition is provided including a derived extracellular matrix having incorporated therein a macromolecular network of polycarboxylate or polyamine macromolecules that have been cross-linked via dihydroxyphenyl linkages. The macromolecular network is interlocked within the extracellular matrix.

[0006] A method of making an implantable composition is also provided. The method includes impregnating a derived extracellular matrix with hyaluronan macromolecules that have hydroxyphenyl side groups substituted thereon; and thereafter reacting the hydroxyphenyl side groups to form dihydroxyphenyl linkages, thereby incorporating a cross-linked macromolecular network of hyaluronan that is interlocked within the extracellular matrix.

[0007] A method of reinforcing a tissue repair, or repairing a tissue defect or gap, is also provided. The method includes providing a patch comprising a derived extracellular matrix having impregnated therein polycarboxylate or polyamine macromolecules that have hydroxyphenyl side groups substituted thereon; reacting the hydroxyphenyl side groups to form dihydroxyphenyl linkages, thereby incorporating a cross-linked macromolecular network of said polycarboxylate or polyamine macromolecules that is interlocked within the extracellular matrix; identifying an appropriate tissue defect or gap in animal or human tissue in vivo; and applying the patch to either reinforce the tissue repair or cover the tissue defect or gap..

[0008] A patch for reinforcing a tissue repair, or repairing a tissue defect or gap, is also provided. The patch includes a derived extracellular matrix

having impregnated therein a macromolecular network of polycarboxylate or polyamine macromolecules that have been cross-linked via dihydroxyphenyl linkages. The macromolecular network is incorporated in and interlocked within the extracellular matrix. In preferred embodiments, the derived extracellular matrix is fascia lata and the macromolecules are hyaluronan molecules that have been substituted with tyramine and cross-linked via dityramine linkages.

BRIEF DESCRIPTION OF THE DRAWINGS

[0009] FIG. 1 is a schematic diagram of an extracellular matrix patch applied over a tendon-to-bone repair site to reinforce the repair.

[0010] FIG. 2 is a structural formula of a hyaluronan molecule.

[0011] FIG. 3 is a schematic diagram of one embodiment of a dihydroxyphenyl cross-linked macromolecular network.

[0012] FIG. 4 is a schematic diagram of an extracellular matrix patch applied to repair a tissue defect between tendon and bone, bridging the gap between the tendon and bone in what would otherwise be an irreparable tear.

[0013] FIG. 5 is a series of two photographs showing the distribution of

HA in fascia ECM (A) treated with water and (B) treated with tyramine-substituted HA followed by cross-linking, as described in Example 1.

[0014] FIG. 6 is a series of three photographs showing representative histologic sections of fascia extracellular matrix grafts following implantation in the rat abdominal wall model for four weeks (hematoxylin & eosin staining, 10X), wherein the fascia was (A) untreated, (B) HA-treated, and (C) TS-HA treated followed by cross-linking, as described in Example 4.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION [0015] Example embodiments are described below and illustrated in the drawings. These embodiments are not intended to be limitations. For example, one or more aspects can be utilized in other embodiments and even other devices or methods.

[0016] Currently, extracellular matrix ("ECM") patches are being investigated for rotator-cuff repair based on the rationale that they can provide a natural and mechanically robust scaffold that reinforces the repair, guides host cell infiltration, and fosters formation of a functional tendon-bone bridge. Extracellular

matrix is the naturally-occurring extracellular part of animal or human tissue that can provide structural support to cells in addition to performing various other functions. According to preferred embodiments, patches and methods disclosed herein utilize human- or animal-derived ECM that is processed and optionally cut into a patch, and augmented ex vivo to incorporate and interlock a cross-linked macromolecular matrix to impart beneficial properties. More generally, as used herein a human- or animal- derived ECM or simply 'derived ECM' is an extracellular matrix material that is excised and harvested from an animal or human, and then processed with or without additional chemical or other treatments ex vivo to provide a patch that can be used to repair (or reinforce the repair of) a tissue defect or gap in an animal or human in need of such repair. The animal or human host who is the source of the derived ECM may or may not be the patient who is to receive the patch, made from derived ECM, to repair a tissue defect or gap.

[0017] Derived ECM may fall into two broad categories: viable derived

ECM and non-viable derived ECM. Herein, viable derived ECM refers to human- or animal-derived ECM as described above, which has not been treated to remove or kill living cells present or suspended in the ECM. Accordingly, viable derived ECM will retain living cells from the host of the ECM, which were present therein upon harvesting. Viable derived ECM may be the preferred choice to prepare an autograft patch for auto-transplantation; i.e. if the tissue defect or gap in need of repair is in the host from which the ECM was derived. Alternatively, viable derived ECM also may be preferred in the case of a derived ECM host who has been tissue-typed and matched with the patient in need of the repair, to provide an allograft patch for allotransplantation . Viable derived ECM may be derived from a living host, and used immediately (optionally even intra-operatively) to repair a tissue defect or gap, either in the ECM host or in another subject. Alternatively, the viable derived ECM may be tissue-typed and banked until needed, depending on the capability to preserve it in a viable state. In a further alternative, viable derived ECM may be derived (e.g. donated) from a recently-deceased host, wherein the cells present in the harvested ECM have not yet died.

[0018] Herein, non-viable derived ECM refers to human- or animal- derived ECM as described above, which has either been treated to remove or kill living cells present when harvested from a viable or recently deceased host, or to remove the components of necrotic cells present when harvested from cadaveric

tissues after a period of time following death. Non-viable derived ECM may be the preferred option in circumstances where there is risk of inducing an unacceptable or undesirable immune response in the recipient of a patch made with the derived ECM, based on the presence of foreign living cells. Alternatively, non-viable derived ECM may be preferred when it is desirable to incorporate specific cells of known source and concentration into the patch to be formed from the derived ECM, for example autograft cells. Non-viable derived ECM also may be preferred because it is generally easier to obtain compared to viable derived ECM. In addition to obtaining non-viable ECM from cadaveric humans or animals, it may also be prepared by treating otherwise viable derived ECM with appropriate agents or techniques to kill and remove the present living cells prior to use.

[0019] The patches described herein, made from or substantially from derived ECM, are applied to repair a tissue defect or gap, or to reinforce such a repair. As used herein, a tissue defect or gap is a space between tissues of a patient (animal or human), or between adjacent portions of the same tissue, which it is desirable to join together. A tissue defect or gap can be formed from a tear or severance of tissues that once were joined, such as when a tendon tears or becomes detached from the associated bone, or when a perforation is formed (either traumatically or through a disease or other natural process) in tissue, such as an ulcer.

[0020] Herein, patches made from derived ECM and methods of using them to repair or reinforce the repair of a tissue defect or gap are disclosed. Fascia lata is one form of derived ECM useful to prepare the patches described herein. Fascia lata can be human- or animal-derived. It is particularly interesting for applications described herein, especially to reinforce tendon-to-bone repairs, because its chemical, structural and material properties are similar to tendon. Fascia lata thus has good tensile strengh, an important property in retaining torn or perforated tissues together. In preferred embodiments, a patch used to reinforce or repair a tissue defect or gap is composed of, or primarily of, fascia lata. When used to repair defects in humans, the fascia lata preferably is human fascia lata, which is readily available as allograft tissue from tissue banks. Alternatively, fascia lata from other animals may be used in humans, including fascia lata derived from, e.g., pigs or cows.

[0021] FIG. 1 schematically illustrates an ECM patch, such as a fascia lata patch as preferred herein, applied and sewn over a tendon-to-bone repair site. In the illustrated embodiment, the tendon is initially repaired directly to the bone from which it has torn. This can be done by conventional techniques, as by suturing. Subsequent to this initial repair, the patch made from derived ECM is applied over the tendon-to-bone repair, and is separately attached to intact tissues located at opposite sides of the repair site. In this manner, the patch provides additional reinforcement to the repair and diminishes the tensile forces experienced by the tendon-to-bone sutures at the repair site. Thus, it will be appreciated that with the use of an appropriate ECM patch and surgical method of fixation, the sutures at the repair site (directly attached to the tendon and bone) may be less prone to tear through either the tendon or the bone. As a consequence, the tendon remains opposed to the bone in a manner that supports natural healing and the formation of a functional tendon-bone bridge. Because fascia lata ECM has mechanical properties similar to tendon, it has the inherent material properties to withstand the physiologic forces experienced during tendon-bone healing.

[0022] In addition to reinforcing the repair site, it is desirable to create and promote an environment conducive to rapid healing, which should reduce the amount of time the sutures must sustain the repair, and in turn reduce the re-tear rate. An environment that inhibits the migration of inflammatory cells and induces the migration of non-inflammatory cells would be beneficial. High molecular weight hyaluronic acid, also known as hyaluronan and abbreviated HA, the basic structure of which is shown in FIG. 2, is a ubiquitous biomolecule that is known to play a critical role in morphogenesis with myriad functions. During both morphogenesis and fetal wound healing, the persistently high levels of high molecular weight HA are thought to promote an intermediate state of cellular adhesion that allows cell shape changes and motility and/or maintenance in the undifferentiated state. Fetal wounds have abundant HA, minimal inflammation, and heal without scar formation. Based on these favorable qualities, HA has been widely investigated in tissue repair strategies. High molecular weight HA (>250 kDa) is known to possess anti-inflammatory properties leading to cell quiescence and decreased fibrosis/scar formation. High molecular weight HA may modulate inflammation by direct interaction with inflammatory cells or by creating a viscous environment that limits inflammatory cell migration and diffusion of cytokines. However, when degraded into small molecular

weight fragments, HA has been shown to be pro-inflammatory and angiogenic, stimulating cell migration, cytokine production and fibrotic tissue formation. Hence in terms of therapeutic efficacy, high molecular weight HA may be beneficial as both an anti-inflammatory agent and a protective barrier against inflammatory byproducts.

[0023] According to an emodiment, high-molecular weight HA (e.g. ~1

MDa) is incorporated into derived ECM to augment the ECM and provide an improved patch compared to using the derived ECM alone. The ECM can be any suitable derived ECM. Preferably, it is fascia lata ECM, and in the case of repairing a tissue defect in a human, preferably human fascia lata ECM. Alternatively, the derived ECM can be, e.g., fascia lata from other animals, or it can be other types of ECM, such as dermis ECM, small intestinal submucosal ECM, pericardium ECM, etc. Still further it can be a combination of derived ECM materials. In preferred embodiments, high-molecular weight HA is incorporated into the derived ECM to prepare an improved patch product. The HA is preferably hydroxyphenyl-substituted hyaluronan, more preferably tyramine-substituted hyaluronan (TS-HA), which can be first impregnated in the ECM, and then immobilized within the ECM by cross-linking of the adducts, e.g. the hydroxyphenyl or tyramine adducts, to form cross-links, e.g. dihydroxyphenyl or dityramine linkages, thereby producing a cross-linked HA macromolecular network, as shown schematically in FIG. 3, incorporated into the derived ECM. By cross-linking in situ within the ECM, the impregnated HA becomes entangled within the ECM and the resulting cross-linked HA macromolecular network becomes entrapped and interlocked therein. In addition, in situ cross-linking allows for the potential of direct cross-linking of HA bound adducts, e.g. tyramine, with tyrosine residues on native proteins within the derived ECM, to covalently bind the HA network to these proteins. Such proteins may be naturally occuring and retained in the derived ECM, or they can be separately incorporated to achieve a desired therapeutic effect. It has been shown that when approximately 5% of the dissacharides present in hyaluronan macromolecules are substituted with tyramine, this allows TS-HA to be discretely cross-linked (or gelled). The structure, and presumably the biological activity, of the cross-linked TS-HA is maintained, as evidenced by the positive binding of HA binding protein, as shown in FIG 5 and described in more detail in the examples below.

[0024] Our data suggest that fascia lata ECM can be enriched with incorporated TS-HA by at least an order of magnitude compared to natural HA

already present in fascia lata ECM, using diffusion methods. Moreover, TS-HA- enhched fascia has been shown to possess increased cell density compared to untreated fascia in a rat abdominal wall model (described in examples below). Data from the examples also suggest that fascia enriched with immobilized TS-HA is associated with less chronic inflammation than untreated fascia in the rat model. Together, these results indicate that incorporating a small concentration (1 -2% tissue weight based on ECM) of high molecular weight HA into fascia lata ECM creates a milieu that enhances cellular infiltration and modulates inflammation - conditions favorable for graft incorporation and perhaps not unlike those found in fetal development. It is to be noted that up to -1 -2% loading (based on initial fascia tissue mass) has been obtained via the diffusion method mentioned above and described more fully hereinbelow. However, it is contemplated that higher TS-HA loading rates may be obtained through more aggressive loading techniques, for example via vacuum, centrifugation or electroporation. The above effects, combined with the tensile properties of fascia lata, support the use of fascia enriched with TS-HA for tendon-to-bone repair in general, and rotator cuff repair in particular. We expect that high-molecular-weight HA will reduce inflammation and promote cell migration into fascia, which would promote integration of the fascia patch with the underlying host tissues. As such the fascia patch would provide reinforcement to the tendon-to-bone repair during healing. This would diminish the occurrence of re-tearing following repair. The foregoing effects of combining HA with fascia lata to reinforce a tendon- to-bone repair will be particularly useful in rotator-cuff repair because of the high tensile forces seen in the shoulder joint during post-operative activites and the poor rates of biologic healing.

[0025] It is believed that HA-enrichment will enhance the host response to fascia such that augmenting rotator cuff repairs with an HA-enriched fascia patch will improve repair outcomes over conventional suture-only repair or augmentation with a patch comprised of only native fascia. It is further believed that enriching fascia ECM with exogenous high molecular weight hyaluronan (HA) - a molecule known to play a critical role in morphogenesis, wound repair, and modulating inflammation - may improve the fascia's ability to foster the formation of a functional tendon-bone bridge in rotator cuff repair by enhancing cellular infiltration and modulating inflammation.

[0026] Exogenous HA may be combined with fascia lata to produce an improved fascia lata ECM material. Fascia lata may be obtained, for example, from tissue banks as described above, as well as other sources. Fascia lata is typically supplied in decellulahzed form as a non-viable ECM material. It is a very dense material, and can be loaded with HA at -1 -2% of the total initial fascia tissue mass using diffusion methods. The added HA is distributed throughout the depth of the fascia scaffold, primarily around large fascicle bundles. As mentioned above, it may be possible to achieve greater loading via alternative and more aggressive techniques, including vacuum, centhfugation or electroporation.

[0027] As mentioned above, fascia lata often will be obtained from tissue banks and other commercial sources. These sources may apply special or even proprietary processing techniques to process the fascia lata before it is used. All of these techniques may have the effect of making the allograft fascia lata seem modestly foreign to the implantee's immune system. It is believed that incorporating small amounts of high-molecular-weight HA into the fascia lata (-1 -2% of the total mass of ECM) will mask degraded or damaged proteins arising from processing and thus reduce the immune response of the implantee and enhance integration of the patch.

[0028] High-molecular weight non-cross-linked HA rapidly diffuses out of the fascia in vitro (e.g. within 1-72 hours). If HA diffuses rapidly out of the fascia, it would not be present to facilitate the beneficial effects described above in an implanted fascia lata (or other ECM) patch during the healing process. Therefore, the HA impregnated in the ECM is augmented to entrap it therein so that it will remain in the patch for at least a period of time during the implantee's healing process. Cross- linked, high-molecular-weight HA diffuses out of fascia lata ECM more slowly than uncross-linked HA of corresponding molecular weight. A cross-linked HA material that has been found particularly useful in this application is prepared by substituting tyramine moieties onto the HA chains and then linking tyramines to form dityramine linkages between HA chains, thus cross-linking or gelling the impregnated HA molecules to entrap and incorporate them into the fascia matrix. The preferred dityramine-cross-l inked HA composition and chemistry is disclosed in U.S. Patents Nos. 6,982,298, 7,368,502, and 7,465,766, the contents of all of which are incorporated herein by reference in their entirety.

[0029] Briefly, in accordance with the incorporated publications, a cross-linked HA or other polycarboxylate or polyamine macromolecular network can be prepared through covalent coupling of hydroxyphenyl-containing compounds, including but not limited to tyramine, through their primary amine groups to carboxyl groups on the long-chain macromolecules via a carbodiimide-mediated reaction. The hydroxyphenyl groups can be added to the macromolecules periodically or randomly along their length via a chemical reaction. Herein, the long-chain macromolecule preferably is HA, which has periodic carboxylic acid groups along its length and is preferred to facilitate a number of desirable effects (described above) beyond the use of un-augmented, derived ECM alone as a patch used to repair tissue defects. The hydroxyphenyl groups are provided as part of smaller molecules having primary amine groups that can be attached to the carboxyl carbon atoms of a carboxylic acid group on the HA macromolecules via the carbodiimide pathway. The reactions are described in detail in the publications incorporated above.

[0030] When substituting onto a HA molecule, suitable hydroxyphenyl- containing compounds include those having a free primary amine that can be used to modify scaffold materials having multiple or periodic CO2H groups, including tyrosine (2-amino-3-(4-hydroxyphenyl)propionic acid) and tyramine (tyrosamine or 2- (4-hydroxyphenyl)ethylamine).

[0031] The second step in preparing the cross-linked HA network is to link the hydroxyphenyl-substituted macromolecules via a dihydroxyphenyl linking structure. In this step hydroxyphenyl groups attached to different HA molecules are linked using a peroxide reagent in the presence of a peroxidase. Of note, some dihydroxyphenyl linking may occur between different hydroxyphenyl groups attached to the same molecule. Peroxidase in the presence of a dilute peroxide (preferably H2O2) is able to extract the phenolic hydroxyl hydrogen atom from hydroxyphenyl containing compounds (such as tyramine) leaving the phenolic hydroxyl oxygen with a single unshared electron, an extremely reactive free radical. The free radical isomerizes to one of the two equivalent ortho-position carbons and then two such structures dimerize to form a covalent bond effectively cross-linking the structures, which after enolizing generates a dihydroxyphenyl dimer, e.g. a dihydroxyphenyl linkage such as a dityramine linkage. Suitable peroxides include hydrogren peroxide. The peroxidase preferably is horseradish peroxidase (HRP). Alternatively, any other suitable enzyme or other agent can be used that is capable of generating free-

radicals for cross-linking long-chain macromolecules that contain hydroxyphenyl groups. Considering the peroxidase enzyme in more detail, the peroxidase can either form hydroxyphenyl radicals required for cross-linking through interaction of hydroxyphenyl groups at the enzyme active site to directly create the desired radicals, or through generation of superoxide radicals, which then diffuse from the enzyme and interact with hydroxyphenyl groups to generate the desired radicals. Other compounds that have the potential to produce the same effect include any porphyrin containing compound, which includes the peroxidase family, hemoproteins, or the structurally related chlorin compounds. A number of other free radical initiators can also be used to crosslink the hydroxyphenyl-modified long-chain macromolecules, as described in detail in the publications incorporated above.

[0032] Returning to the augmented, derived ECM patches disclosed here, preferably the high-molecular-weight HA molecules to be incorporated into fascia lata or other suitable derived ECM have an average molecular weight of 250 kDa or greater, more preferably 500, 800 or 900 KDa or greater, more preferably 1 MDa or greater. Preferably, the tyramine-substitution rate on the HA molecules is about or less than five percent based on available substitution sites as disclosed in the aforementioned publications. It is desirable that tyramine-substitution onto the HA molecules be completed prior to impregnating the derived ECM. The substituted HA macromolecules of the desired molecular weight (preferably greater than 250 kDa as noted above) are then impregnated into the derived ECM via a suitable technique, such as passive diffusion as noted above. Subsequently, once the desired degree of TS-HA loading has been achieved (e.g. -1-2%), the TS-HA- impregnated ECM is submerged in a solution including dilute hydrogen peroxide and the appropriate enzyme, e.g. HRP, to promote and facilitate the cross-linking reaction to generate dihydroxyphenyl (e.g. dityramine) linkages, to thereby interlock the impregnated HA macromolecules to produce an interlocked macromolecular network within the derived ECM. The submersion described here can be done either once the TS-HA-impregnated patch has been cut to the desired dimensions for the intended repair, or it can be done using a larger sheet of the material, which can be subsequently cut to size as needed for particular repairs.

[0033] The low tyramine-substitution rates described above allow HA to be discretely cross-linked while maintaining the majority of the HA molecules in their native conformation as evidenced by positive binding of hyaluronan binding protein

(FIG. 5 described in the examples below). In addition to preserving the HA within the derived ECM, cross-linking as described above may inhibit or suppress the breakdown of HA into molecules of relatively lower molecular weight.

[0034] Thus far, the HA-augmented, derived ECM patch has been described, in conjunction with FIG. 1 , as being applied over an extant surgical repair to attach or re-attach adjacent torn or separated tissues. As already explained, in this embodiment the patch serves to reinforce the surgical repair directly at the defect, with additional benefits realized as a result of incorporating the cross-linked macromolecular HA network in the patch. In an alternative embodiment, the patch disclosed herein also may be used as a bridging material in the case where the gap between separated or torn tissues, such as a tendon and the associated bone, is too large to repair conventionally, e.g. via direct suturing of the tissues. This can happen, for example, when there is no longer sufficient native tissue to close the tear or gap to bring opposed portions into contact for suturing. In one embodiment shown schematically in FIG. 4, a derived ECM patch, preferably augmented with impregnated and cross-linked HA as described above, is incorporated at the bone- tendon interface and fixed to both at its respective ends to bridge a gap that is otherwise too broad to be repaired conventionally. In this embodiment, the derived ECM patch will itself be integrated to both the bone and the tendon to act as a surrogate tendon-bone bridge, essentially repairing what traditionally would have been an irreparable gap or tear. Because fascia lata has properties very similar to tendon, it is already well suited to function in this role as the derived ECM of the patch. In addition, the entrapped (cross-linked) HA may promote integration of fascia lata to the bone and/or the tendon for reasons already discussed, which would facilitate the formation of a functional bridge.

[0035] While HA-impregnated fascia lata ECM has been described primarily with respect to rotator-cuff repairs, it will be appreciated that similar methods and materials as described here could also be adapted to other tendon-to- bone repairs, repairs of other connective tissues (such as ligaments or ligament-to- bone), soft-tissue repairs such as repair of lacerated or transferred muscles, as well as other repairs where a tissue defect or gap has occurred either through injury (such as a hernia) or surgically. In these cases, an HA-impregnated and incorporated (via cross-linking) derived ECM matrix, such as fascia lata, which can be prepared and supplied in the form of a patch having appropriate dimensions, can

be applied over the defect (e.g. tear) either to form a tissue bridge that can be incorporated and remodeled via the implantee's own healing response into appropriate tissue to complete the repair, or to reinforce a surgical repair between the adjacent tissues while it heals. The presence of the high molecular weight HA trapped in the derived ECM should enhance the implantee's remodeling response to incorporate the derived ECM (e.g. fascia) into native tissues, and should suppress an inflammatory response that may interfere with healing.

[0036] While the derived ECM patch having impregnated and cross- linked HA can reinforce a repair site, the repair still may be prone to failure before biological healing occurs due to the sutures tearing through either the tendon or the bone, or through the ECM material (such as fascia lata) itself. Therefore, additional steps to induce functional integration at the repair site as quickly as possible, independent of sutures, are also desirable. Application of these steps may be desirable for repairs as depicted schematically in FIG. 1 and FIG. 4. For this purpose, it is believed that a gel made of tyramine-cross-l inked collagen would be useful, for example, if injected between the ECM patch and the underlying tissue to be repaired, or between the implantee's tendon and bone or other separated tissue at the repair site, while suturing or after suturing is complete. Like HA, collagen is a polycarboxylate macromolecule capable of being substituted with tyramine or other hydroxyphenyl groups to facilitate dityramine (dihydroxyphenyl) linkage. Appropriate chemistry is described in detail in the publications incorporated above. The form of the collagen to be substituted with tyramine ranges from highly purified recombinant collagen to collagen that is a component of a tissue extract such as gelatin. To prepare a collagen gel suitable for this application, preferably collagen having an average molecular weight greater than 50 kDa, more preferably 100 kDa, is used. In this embodiment, tyramine-substituted collagen ("TS-collagen") is preferably cross- linked in situ between the opposed, sutured tissues allowing the TS-collagen to penetrate the opposed tissues as it cross-links. The TS-collagen would then crosslink (or gel) around the existing tissues and potentially cross-link to existing tyrosine residues within the tissues in a manner similar to the integration of TS-HA with proteins that may be present within fascia lata ECM, or within other derived ECM in the patch as described above. Use of TS-collagen in this manner may optimize the integration of the opposed tissues, serving as both a glue and potentially an inductive agent for repair.

[0037] While cross-linking of both HA and collagen described herein is preferably achieved via dityramine linkages as mentioned above and more fully described in the incorporated publications, it is contemplated that other dihydroxyphenyl linkages also may be used, e.g. via substitution and carbodiimide- mediated linking of suitable hydroxyphenyl adducts onto the HA or collagen chains. This is also further described in the incorporated publications. Direct incorporation of cells, biologies, and other active ingredients (such as drugs or other therapeutic compounds) into the TS-HA or TS-collagen macromolecular networks to induce healing is also contemplated as described in the incorporated publications.

[0038] While the present methods and patch are described primarily with respect to augmenting fascia lata ECM as the derived ECM as disclosed herein, it is to be recognized that the methods and compositions described herein may be applied analogously to incorporate TS-HA into other ECM materials besides fascia, such as dermis, small intestine submucosa, pericardium, or even other scaffold materials derived from polymeric biomaterials (i.e. animal-derived materials). In addition, the foregoing methods have been disclosed primarily in connection with impregnating and incorporating into the derived ECM HA macromolecules. The HA molecules are cross-linked by linking hydroxyphenyl, preferably tyramine, adducts attached to those molecules to generate an interlocked HA network that is retained in the derived ECM matrix. The resulting cross-linked ECM matrix accordingly diffuses out of the matrix more slowly, increasing the length of time during healing when it is present to produce beneficial effects. It is contemplated that other polycarboxylate or polyamine macromolecules other than HA, which lend themselves to substitution with hydroxyphenyl adducts followed by linkage via the above- described chemistry, may be incorporated into a derived ECM patch, such as a fascia lata patch, via analogous methods as disclosed for HA to produce corresponding desirable or therapeutic effects associated with those other macromolecules. Alternative (other than HA) polycarboxylate or polyamine macromolecules that may be impregnated into a dehved-ECM matrix and then cross-linked via linkage of hydroxyphenyl adducts thereon may include molecules of or based on other glycosaminoglycans (such as heparin) and proteins (such as collagens). Such other polycarboxylate or polyamine macromolecules, which provide or may provide associated desirable effects, can be incorporated into the derived- ECM matrix via similar methods disclosed for HA herein and in the incorporated

publications. In a further alternative, macromolecules already possessing native hydroxyphenyl side groups may be incorporated into a derived ECM matrix and then cross-linked to form dihydroxyphenyl linkages between native hydroxyphenyl side groups. In this case, the step of substituting hydroxyphenyl-containing groups onto the macromolecule can be omitted, because such groups are already present.

[0039] In summary, in a broad aspect there is disclosed a derived ECM matrix that incorporates a cross-linked macromolecular network interlocked within the derived ECM, which network includes polycarboxylate or polyamine macromolecules that have been substituted with hydroxyphenyl adducts and cross- linked via dihydroxyphenyl linkages. In a preferred aspect, the ECM matrix is fascia lata, preferably human fascia lata ECM, and the hydroxyphenyl-substituted polycarboxylate or polyamine molecules are tyramine-substituted HA molecules that have been cross-linked via dityramine linkages, to interlock a macromolecular HA network within the derived fascia lata matrix. The resulting matrix is cut or formed into dimensions suitable to be used as a patch, either to reinforce the surgical repair of a tissue defect or gap, or to repair the defect as in the case of a gap that is too broad for conventional suture-based repair. A patch composed of derived fascia lata ECM incorporating HA in the form of a macromolecular network that is cross-linked via dityramine linkages and thereby mechanically interlocked and retained within the ECM is contemplated to be particularly useful in repairing rotator-cuff tears, as well as other tendon-to-bone or connective tissue repairs.

[0040] Still further aspects and features of the invention will be apparent from the following examples, which are provided by way of illustration and not limitation.

[0041] Example 1

[0042] Acellularized human fascia lata was obtained from the

Musculoskeletal Transplant Foundation (Edison, NJ). Prior to shipment the fascia was processed according to a published procedure, as disclosed in WO/2006/101885, publication date of September 28, 2006, which is incorporated by reference, with modifications, hereafter referred to as "standard pre-processing". Specifically, the modified process used by the Musculoskeletal Transplant Foundation comprised the following steps: (1 ) isolating fascia from a suitable donor; (2) processing the fascia including inspection for visual defects, trimming and soaking the tissue in phosphate buffered saline (PBS) and rinsing same with sterile

PBS; (3) soaking the tissue in an antibiotic composition and rinsing same multiple times in sterile PBS; (4) processing the tissue by cutting the tissue to size; and (5) freezing the tissue in sterile PBS at -20 ⁰ C and shipping on dry ice.

[0043] HA was impregnated in fascia lata as follows. Approximately 4x4 cm samples of fascia lata were lyophilized for at least 24 hours. Following lyophilization, individual fascia samples were rehydrated in a 0.75% (w/v) TS-HA solution (corresponding to mass-volume percentage, such that the solution includes 0.75 g TS-HA per 100 ml) (~1.5 ml/cm ² ) for 24 hours at 37 ⁰ C on a shaker. Molecular weight of TS-HA used was 900 kDa - 1 MDa. The substitution rate of tyramine was -5%. After impregnation of fascia with TS-HA, fascia was rinsed of excess TS-HA with high purity water for -30 seconds and blotted two times per side. To cross-link the TS-HA, fascia was submerged into a 0.3% hydrogen peroxide solution for -30 seconds and incubated overnight at 4 ⁰ C to allow the reaction to continue. Subsequently, fascia was rinsed of excess hydrogen peroxide with high purity water for -30 seconds and lyophilized. Non-cross-linked samples were submerged in high purity water for -30 seconds, incubated overnight at 4°C, rinsed, and lyophilized. Water controls were treated by incubation in high-purity water for 24 hours at 37°C on a shaker, subjected to hydrogen peroxide, and subsequently lyophilized.

[0044] Fluorophore assisted carbohydrate electrophoresis (FACE) was used to quantify the HA content of the treated fascia samples, according to Calabro A. et al., Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by fluorophore-assisted carbohydrate electrophoresis (FACE), Glycobiology 2000, 10(3), pp. 273-81 ; and Calabro A., Adaptation of FACE methodology for microanalysis of total hyaluronan and chondroitin sulfate composition from cartilage, Glycobiology 2000, 10(3): 283-293. Results are shown in TABLE 1.

TABLE 1.

Fluorophore-assisted carbohydrate electrophoresis analysis (FACE) of HA content in diffusion-treated 4x4 cm fascia (STANDARD PRE-PROCESSING).

[0045] According to an alternative approach, fascia lata was preprocessed according to the method of the WO/2006/101885 publication, incorporated above (i.e., standard pre-processing), except that samples of fascia lata were 1x1 cm. Diffusion treatments of the fascia lata were evaluated based on TS-HA solutions that were 0.5, 0.75, or 1 % (w/v). FACE was used to quantify the HA content of the treated fascia samples, according to the above-mentioned Calabro publications. Results are shown in TABLE 2.

TABLE 2.

Fluorophore-assisted carbohydrate electrophoresis analysis (FACE) of HA content in diffusion-treated 1x1 cm fascia (STANDARD PRE-PROCESSING).

[0046] According to another alternative approach, fascia lata was preprocessed according to the method of the WO/2006/101885 publication,

incorporated above, with modifications. Specifically, bulk fascia lata was prepared and shipped (hereafter referred to as "bulk pre-processing"). Bulk fascia is fascia that is processed according to WO/2006/101885, except that tissue is subjected to only 4 hours of the antibiotic soak rather than the full 24 hours. Upon receipt of shipped bulk fascia lata, the antibiotic soak process was continued in house by subjecting the fascia to an additional 20 hours soak prior to freezing and lyophilization. Additionally, bulk fascia was cleaned in house of excess loose connective tissue and fat. Bulk fascia lata (-4x4 cm) was then treated using 0.75% (w/v) TS-HA solution. FACE was used to quantify the HA content of the treated fascia samples, according to the above-mentioned Calabro publications. Results are shown in TABLE 3.

TABLE 3.

Fluorophore-assisted carbohydrate electrophoresis analysis (FACE) of HA content in diffusion-treated 4x4 cm fascia (BULK PRE-PROCESSING).

[0047] According to a further alternative approach, fascia lata was preprocessed according to the method of the WO/2006/101885 publication, incorporated above, with modifications. Specifically, bulk fascia that was not subjected to lyophilization prior to TS-HA impregnation was treated (hereafter referred to as "bulk, not lyophilized, pre-processing"). Instead, fascia was kept hydrated in PBS at 4 ⁰ C until subjected to TS-HA treatment. Fascia (-4x4 cm) was treated using 0.75% (w/v) and 2% (w/v) TS-HA solutions. FACE was used to quantify the HA content of the treated fascia samples, according to the above-mentioned Calabro publications. Results are shown in TABLE 4.

TABLE 4.

Fluorophore-assisted carbohydrate electrophoresis analysis (FACE) of HA content in diffusion-treated 4x4 cm fascia (BULK, NOT LYOPHILIZED, PRE-PROCESSING).

[0048] In an additional analysis method, TS-HA treated fascia samples

(1x1 cm, diffusion, standard pre-processing) were rehydrated in 0.2 ml of saline for 5 minutes and embedded in Tissue-Tek(R) OCT compound. Five micron longitudinal frozen sections were cut and stained with biotinylated HA binding protein (HABP, Calbiochem, San Diego, CA) and Alexa Fluor(R) 488 conjugated streptavidin (Molecular Probes, Carlsbad, CA) to visualize TS-HA distribution in the tissue. HABP staining demonstrated TS-HA incorporation throughout the depth of the tissue, primarily around large fascicle bundles (FIG. 5). The arrow in FIG. 5B denotes staining of TS-HA. It is desirable that the added HA is distributed throughout the fascia, rather than being layered in high concentration on the surface, to promote more uniform benefits of HA incorporation, such as suppression of an inflammatory response. Further, the binding of HABP demonstrates that the TS-HA remains biologically active after cross-linking.

[0049] Example 2

[0050] A method for impregnation of fascia lata with HA based on vacuum is as follows. Fascia lata, pre-processed according to either standard or bulk methods as described in Example 1 above, was lyophilized for at least 24 hours. Following lyophilization, fascia was cut into 3 cm diameter samples and rehydrated in high purity water for -10 minutes. Individual fascia lata samples were then mounted onto the platform of a modified Millipore Steriflip® filtration system. The top vessel was filled with -10 ml of TS-HA solution and secured to the filtration system. Vacuum was applied and the TS-HA solution was pulled through the fascia lata. This

was repeated to increase the number of passes of solution through the fascia lata. The concentrations evaluated were 0.5, 0.75, 1 % (w/v), and increasing concentrations. FACE was used to quantify the HA content of the treated fascia samples, according to the above-mentioned Calabro publications. Results are shown in TABLE 5.

TABLE 5.

Fluorophore-assisted carbohydrate electrophoresis analysis (FACE) of HA content in vacuum-treated 3 cm diameter fascia.

[0051] Example 3

[0052] A method for impregnation of fascia lata with HA based on centrifugation is as follows. Fascia lata, pre-processed according to the standard method as described in Example 1 above, was received lyophilized, cut into 1 cm diameter samples, and rehydrated in high purity water for -10 minutes. Individual fascia samples were placed into a centrifugal filter device corresponding to a Millipore Centhcon® tube and covered with 1.5 ml of HA (for this example, not TS- HA) solution. The Centhcon tube was then centrifuged at 500Og for ~1 hour. The solution "filtrate" was placed back into the Centhcon tube and the centrifugation was repeated once more. The concentration evaluated was increasing concentrations (0.05→0.1 →0.2→0.5% (w/v)). To visualize HA distribution in fascia samples, the samples were rehydrated in 0.2 ml of saline for 5 minutes and embedded in Tissue-

Tek® OCT compound. Five micron longitudinal frozen sections were cut and stained with biotinylated HA binding protein (HABP, Calbiochem, San Diego, CA) and Alexa Fluor(R) 488 conjugated streptavidin (Molecular Probes, Carlsbad, CA). HABP staining demonstrated HA incorporation throughout the depth of the tissue, similar to FIG. 5.

[0053] Example 4

[0054] An experiment was conducted to evaluate the host response to

TS-HA enriched fascia ECM in a rat abdominal wall model.

[0055] Experimental Design. Acellulahzed human fascia lata was cut into 1x1 cm samples and distributed into three treatment groups: TS-HA with cross- linking, HA without cross-linking, or untreated controls. Samples were treated in a similar manner as described in Example 1. At surgery, fascia grafts were rehydrated and implanted into a partial-thickness defect of the anterior sheath of 6 rats, n=2 per group, as described in the following section. Additional details are provided in the following reference: Valentin JE et al., Extracellular matrix bioscaffolds for orthopaedic applications. A comparative histologic study. J. Bone Joint Surg. [Am.] 2006;88(12):2673-86. At four weeks, rats were euthanized and the graft and surrounding muscle were harvested for histology. Cell density within the graft was quantified using image processing techniques detailed below. Inflammation was qualitatively assessed by a board-certified pathologist.

[0056] Rat Abdominal Wall Defect Model. Adult, male Lewis rats (450-

600 gm) were used. Rats were anesthetized and prepared for sterile surgery. Via a ventral midline incision, a partial-thickness 1x1 cm defect was created in the anterior sheath adjacent to the linea alba. The anterior sheath was removed and the underlying rectus muscle, transversalis fascia, and peritoneum were left intact. A 1x1 cm fascia ECM patch was rehydrated in saline for ten minutes and then secured into the defect using four corner sutures of 5-0 Prolene. Securing the fascia theoretically allowed the ECM patch to be subjected to the mechanical forces delivered by the adjacent native abdominal wall musculature. The skin incision was closed and the animals were allowed to recover from anesthesia under a heating lamp before being returned to individual cages for the duration of the study.

[0057] Histologic Processing and Analysis. At euthanasia, the fascia graft and surrounding muscle were harvested, fixed in 4% paraformaldehyde and

processed routinely for paraffin embedding. Five micron thick longitudinal sections of each sample were cut and stained with hematoxylin and eosin (H&E). RGB images of the H&E sections were acquired using a Retiga 2000R CCD digital camera (1600x1200 pixel, Q-lmaging, Burnaby, B.C., Canada) attached to a Leica DM 4000 upright microscope (Heidelberg, Germany) fitted with a 10X objective. Image acquisition was fully automated using an X, Y, Z-motorized stage (Prior Scientific, Rockland, MA) managed by Objective Imaging's Oasis 4i controller (Kansasville, Ml) and Image-Pro 6.0 software (Media Cybernetics, Silver Spring, MD). Images were captured of the entire length and width of each sample and knitted into a single montage. Image-Pro was used to count hematoxylin stained nuclei and determine respective tissue area on three sections per rat. Inflammation was qualitatively rated by a board-certified pathologist on one section per rat.

[0058] Results: Cell Number in Fascia ECM. At four weeks the histologic appearance of the fascia in all groups characteristically exhibited a highly cellular periphery and fewer cells in the central portion of the graft (FIG. 6). The fascia graft and underlying muscle layers are denoted (FIG. 6). Increased cellularity was present within the fascia in the HA treated groups (FIG. 6B-C) compared to untreated fascia (FIG. 6A). Cells were counted from a manually selected region of interest on the montaged images that included the entire fascia graft (~4 mm ² ). The cell densities are shown in TABLE 6. Because of the small sample size in this study, the six measures from the two rats for each group were used for preliminary statistical comparisons using analysis of variance. At four weeks, cell density within TS-HA treated grafts (2859 ± 295 cells/mm ² ) was significantly greater than untreated grafts (2102 ± 545 cells/mm ² ) (p=0.02). Cell density within uncross-l inked HA-treated grafts (2724 ± 662 cells/mm ² ) may also prove to be different from untreated grafts (p=0.06) when assessed in a larger sample size.

TABLE 6.

Cell density within the fascia grafts after 4 weeks implantation in a rat abdominal wall model.

[0059] Results: Chronic Inflammation in Fascia ECM. Slides from each of the six rats were blinded and ranked on two occasions separated by a period of two months by a board-certified pathologist. Slides were ranked for "chronic inflammation" in and around the graft, as defined primarily by the presence of lymphocytes and plasma cells. The rankings are shown in TABLE 7. There was variability in the degree of inflammation observed in the HA treatment without cross- linking group. Based on these rankings it is suggested that TS-HA treatment plus cross-linking is associated with the least chronic inflammation of the three groups. However, it should be noted that the evaluation was limited by an inability to distinguish lymphocytes from macrophages with certainty. Note that Rat 5 was considered to have the highest degree of inflammation (TABLE 7) and the greatest number of cells (TABLE 6) whereas Rat 4 had the least inflammation but still a high number of cells. This outcome suggests that the cell types making up the total cell counts in each group might be different and probably reflects an attempt by the pathologist to distinguish chronic inflammation (lymphocytes and plasma cells) from macrophages and stromal cells in the rankings.

TABLE 7.

Chronic inflammation ranking. Slides were ranked for lymphocytes and plasma cells in and around the fascia graft; Rankings were repeated twice, 1 = most, 6=least.

[0060] Discussion. These data are important because they demonstrate use of the rat abdominal wall model to assess host response to fascia ECM scaffolds. Further, these data suggest that HA treatment will increase the total cell number in fascia ECM compared to untreated fascia. These data also provide the basis for the hypothesis that fascia enriched with TS-HA plus cross-linking will be associated with fewer lymphocytes and plasma cells but a similar number of macrophages and giant cells compared to fascia enriched with TS-HA without cross- linking or fascia treated with water. Finally, these data demonstrate the need for positive identification of cell types in order to definitively interpret the effect of HA treatment on host response.

[0061] Example 5

[0062] An experiment was conducted to evaluate the in vitro release of

TS-HA from TS-HA enriched fascia ECM.

[0063] Experimental Design. Standard pre-processed human fascia lata (4x4 cm) was treated with TS-HA with or without cross-linking using the diffusion methods described in Example 1 (n=2 per experimental group). From each graft, a 0.8x1 cm sample was cut. Fascia samples were rehydrated in 1.2 ml of PBS and incubated for 72 hours at 37 ⁰ C on a shaker. After the 72 hours, fascia pieces were lyophilized and TS-HA content was measured with FACE. The initial TS-HA content of the incubated samples was estimated from the average of several small pieces

sampled over the entire 4x4 cm graft that were not incubated, using FACE according to the above-mentioned Calabro publications.

[0064] Results: In Vitro Release of TS-HA from TS-HA Treated Fascia.

The average percent release of TS-HA after 72 hours in PBS from TS-HA treated fascia with cross-linking was 73% (n=2, with results of 70.1 % and 76.6% release). The average percent release of TS-HA after 72 hours in PBS from TS-HA treated fascia without cross-linking was 81 % (n=2 with, results of 82.8% and 78.5% release).

[0065] Discussion. These data suggest that cross-linking will slow the release of TS-HA from fascia ECM compared to without cross-linking. Additional experiments and a greater sample size are desirable to determine if longer cross- linking time (submersion in peroxide solution with HRP) will further entrap the TS-HA molecules in the fascia ECM and slow their release to an even greater extent.

[0066] Although the above-described embodiments constitute the preferred embodiments, it will be understood that various changes or modifications can be made thereto without departing from the spirit and the scope of the present invention as set forth in the appended claims.