WO2008134491A1 - A biodegradable vascularizing membrane and methods thereof - Google Patents

Abstract

The present invention provides compositions and methods useful for regenerative medicine, biosensors and other implantable applications where vascularization of tissues is required. In particular, the present invention provides biodegradable, polymeric, vascularizing membranes and methods of use thereof that support long-term vascularization within an otherwise avascular fibrous region via the porous microarchitecture of a degrading membrane, without the release of exogenous growth factors.

Description

A BIODEGRADABLE VASCULARIZING MEMBRANE AND METHODS THEREOF

The present application claims priority to U.S. Provisional Patent Application Serial Number 60/926,208, filed April 25, 2007, the disclosure of which is herein incorporated by reference in its entirety.

FIELD OF THE INVENTION

The present invention provides compositions and methods useful for regenerative medicine, biosensors and other implantable applications where vascularization of tissues is required. In particular, the present invention provides biodegradable, polymeric, vascularizing membranes and methods of use thereof that support long-term vascularization within an otherwise avascular fibrous region via the porous micro-architecture of a degrading membrane, without the release of exogenous growth factors.

BACKGROUND OF THE INVENTION

The ability to control the foreign body response to implanted biomaterials is crucial to their function in vivo. The inflammatory response to foreign materials typically leads to cell and extracellular matrix remodeling processes at the interface of the material and host tissue. These processes consist of macrophage adhesion and aggregation at the material surface, formation of a dense collagen capsule that may vary in thickness depending on the inflammatory nature of the material, and deposition of an outer layer that consists of loosely packed vascularized connective tissue (Sieminski and Gooch, 2000, Biomaterials 21 :2232- 2241; Dziubla and Lowman, 2004, J. Biomed. Mater. Res. 68A:603-614). The densely-packed collagenous capsule serves as a barrier to transport by both increasing the resistance to diffusion as well as increasing the distance that nutrients from the outermost vascular layer must travel (Sieminski and Gooch, 2000; Sharkawy et al., 1997, J. Biomed. Mater. Res. 37:401-412). For bio-engineered cellular structures and implantable biosensors, reducing fibrous capsule thickness and maximizing the amount of blood and nutrient supply to the cells is essential. Several methods have been researched to induce angiogenesis in and around implanted biomaterials for this purpose. Non-degradable systems have focused mostly on porosity as the controlling factor for neovascularization (Sieminski and Gooch, 2000; Sharkawy et al., 1997, J. Biomed. Mater. Res. 37:401-412; Brauker et al, 1995, J. Biomed. Mater. Res. 29: 1517-1524), while degradable systems generally incorporate one or more angiogenic growth factors to induce new blood vessel formation (Perets et al., 2003, J. Biomed. Mater. Res. 65A:489-497; Smith et al., 2004, Tissue Eng. 10:63-71; Nomi et al., 2002, MoI. Aspects Med. 23:463-483).

The disadvantage to using non-degradable materials is the tendency for an uncontrolled persistent foreign body response by the host, often times leading to the failure of the implanted material. The use of resorbable materials has a distinct advantage in that as the material degrades, it is replaced by native tissue. By coupling the release of growth factors to material degradation, angiogenesis can be artificially induced. While this method is effective in increasing vessel density at the site of drug release, the effect on remote sites within the body without tightly controlled delivery can be dangerous. Excessive angiogenesis has been linked to the spread of cancer and the growth of tumors in the body (Nomi et al., 2002), although the dangers of exogenous growth factors are theoretical and have not been demonstrated clinically. In addition to these potential health risks, there is also evidence that vessels produced using angiogenic growth factors tend to regress after the stimulus is removed (Ravin et al., 2001, J. Biomed. Mater. Res. 58:313-318) and the combination and temporal delivery of growth factors required for vessel maturation remain under investigation (Richardson et al., 2001, Nat. Biotech. 19: 1029-1034). For these reasons, what are needed are alternative compositions and methods for increasing vessel density around a cell-scaffold construct.

SUMMARY OF THE INVENTION

The present invention provides compositions and methods useful for regenerative medicine, biosensors and other implantable applications where vascularization of tissues is required. In particular, the present invention provides biodegradable, polymeric, vascularizing membranes and methods of use thereof that support long-term vascularization within an otherwise avascular fibrous region via the porous microarchitecture of a degrading membrane, without the release of exogenous growth factors. The present invention provides compositions and methods for the vascularization of the membrane-host tissue interface that is modulated by the porous structure of the biodegradable membrane, which changes as a function of membrane degradation kinetics throughout the vessel maturation process.

In some embodiments, the vascularizing potential of porous membranes is made from poly(L-lactic) acid (PLLA) and poly(DL-lactic-co-glycolic) acid (PLGA) with lactide to glycolide ratios of 75:25 and 50:50 or equivalent materials is demonstrated. For example, PLLA provides slow degradation kinetics (approximately 50% degradation at 1 year) and ability to maintain its porous microarchitecture during vessel formation and maturation. PLGA is useful because its degradation rate is readily increased relative to PLLA without significantly changing the chemistry of components of the membrane (e.g., complete degradation at 8-10 weeks for 50:50 lactide:glycolide ratio). Both materials are, individually, commonly used in tissue engineering applications as their properties are well characterized and they are known to be biocompatible (Gunatillake and Adhikari, 2003, Eur. Cell Mater. 5:1-16).

In some embodiments, the porous membrane comprises a first component that is a polymer having slow degradations kinetics and a second component comprising a polymer that has faster degradation kinetics. In some embodiments, the ratio for the first and second components is from 90: 10 to 50:50, with all intervals therebetween contemplated.

In some embodiments, the present invention provides for polymers with surface erosion biodegradable characteristics. In some embodiments, the polymers comprise a membrane with a pore size of at least 50μm, at least 60 μm, at least 70 μm, at least 80 μm, at least 90 μm, or at least 100 μm. In some embodiments, the polymers comprise a membrane with a porosity at least 50%, at least 60%, or at least 75%. In some embodiments, the polymers of the present invention comprise a biodegradable membrane that encapsulates cells. Cells for incapsulation include, but are not limited to, pancreatic islet cells, stem cells, anti- inflammatory inducing cells, and the like. In some embodiments, the present invention provides a biodegradable membrane that delivers proteins, nucleic acids, and/or therapeutic drugs to a subject. As such, in some embodiments the present invention provides compositions and methods for delivery of therapeutic compounds to a subject and/or tissue is a subject by applying said therapeutic compound (e.g., small molecule, protein or fragment thereof, nucleic acid or fragment thereof, siRNA molecule, drug, etc.) in or on said polymer membrane and implanting said polymer membrane in a subject. In some embodiments, the polymer membranes are further impregnated, or associated with, a test compound and applied to a subject. For example, a test compound could be a potential therapeutic compound for treating cancer or other diseases. In some embodiments, biodegradable polymer membranes are used in combination with test compounds in vitro and in vivo.

In some embodiments, the present invention provides methods for tissue engineering using compositions of the present invention. In some embodiments, methods for tissue engineering include, but are not limited to, the provision of a scaffold for tissue growth, the provision of a membrane capable of vascularization, the provision of a membrane for delivery of therapeutic or test compounds to a tissue, etc. A skilled artisan will recognize other applications for a biodegradable polymer membrane of the present invention.

In some embodiments, the present invention provides compositions comprising a surface erodable biodegradable porous membrane comprising poly (α-hydroxyester) polymers. In some embodiments, the poly (α-hydroxyester) polymers comprise poly(L-lactic acid) and/or poly(DL-lactic-co-glycolic) acid. In some embodiments, the membrane has a pore size of at least 60 microns. In some embodiments, the membrane has a porosity of at least 75%.

In some embodiments, the present invention provides a method of inducing vascularization within a tissue comprising implantation of a biodegradable membrane as described herein in a subject, thereby inducing vascularization within a tissue.

In some embodiments, the present invention provides a method for assessing tissue perfusion of a sample in vitro or in vivo comprising implanting into a tissue a membrane as described herein and applying 4-dimensional elastic light scattering spectroscopy to said tissue there assessing tissue perfusion.

DESCRIPTION OF THE FIGURES

Figure 1 shows A-C) SEM images of PLLA, PLGA 75:25, and PLGA 50:50 membranes upon fabrication, D-F) SEM images of PLLA and PLGA 75:25 membranes after ten weeks, and PLGA 50:50 membrane after 4 weeks of incubation in .01M, pH 7.4 PBS at 37⁰C. Scale bars are lOOμm.

Figure 2 shows in vitro polymer degradation data. PLLA membranes showed little mass loss, while PLGA 50:50 membranes degraded significantly. PLGA 75:25 membranes showed intermediate degradation kinetics. Figure 3 shows H&E images of subcutaneous tissue samples harvested at ten weeks post-implantation. All images shown are from implants harvested from the lower back. A) PLLA membrane, B) PLGA 75:25 membrane, C) PLGA 50:50 membrane, D) THERACYTE membrane. Right hand side images represent enlarged views of the boxed area of interest shown on the left. Arrows in these images indicate blood vessels. FC = fibrous capsule, BV = blood vessel, VM = vascularizing membrane, IM = immunoisolation membrane, WF = woven fibers. Scale bars are lOOμm. Figure 4 shows vascular density in the fibrous capsule within lOOμm of implants at two and ten weeks post implantation. The dashed lines represent the range of vascular density in normal subcutaneous tissue. * Significant difference (p < .01) between the implant material and normal tissue, f Significant difference (p < 0.05) between the implant material and PLGA 50:50 membranes. $ Significant difference (p < 0.01) between the implant and THERACYTE. Figure 5 A and B show immunohistochemistry of total and mature blood vessels within the fibrous capsule at ten weeks. A) CD31 marker (brown) was used to detect endothelial cells, and an α-smooth muscle actin antibody (blue) was used to detect pericytes associated with mature vessels, except for capillaries. Sections shown are ten weeks post implantation. A. PLLA membrane, B) THERACYTE membrane. C and D show single labeling showing only mature vessels at ten weeks. C) PLLA membrane, D) THERACYTE membrane. FC = fibrous capsule, IM = immunoisolating membrane, VM = vascularizing membrane, WF = woven fiber mesh. Scale bars are lOOμm.

Figure 6 shows blood vessel density within lOOμm of the implant edge and within the fibrous capsule obtained from immunohistochemistry for CD31 endothelial cell membrane protein. Dashed lines represent normal tissue density. * Significant difference between implant and normal subcutaneous tissue (p < 0.05). f Significant difference between implant and PLGA 50:50 membranes. J Significant difference between implant and THERACYTE.

Figure 7 shows mature blood vessel density (excluding capillaries) within the fibrous capsule of cross sections that were double labeled for CD31 and α-smooth muscle actin antibodies. Dashed lines represent mature vessel density in normal tissue * Significant difference between implant and normal subcutaneous tissue (p < 0.05). f Significant difference between implant and PLGA 50:50 membranes.

Figure 8 shows A) Average contribution of immature vessels plus capillaries and mature vessels (minus capillaries) to the total blood vessel density at two weeks. B) Contribution of immature vessels plus capillaries and mature vessels (minus capillaries) to the total vessel density at ten weeks. Numbers represent vascular density in the fibrous capsule within lOOμm of the implant. Figure 9 shows EDl staining (brown) of tissue sections two weeks post implantation. Nuclear counterstaining is shown in purple A) PLLA membrane section. B) PLGA 75:25 membrane section. C) PLGA 50:50 membrane section. D) THERACYTE membrane section. FC = fibrous capsule, IM = immunoisolation membrane, VM = vascularizing membrane, WF = woven fibers. Scale bars are lOOμm.

Figure 10 shows macrophage density in the fibrous capsule and within lOOμm of the implants. Dashed lines represent macrophage density in normal subcutaneous tissue. ^Significant difference (p < 0.05) between implant and normal tissue.

Figure 11 shows A) Sample curve of the backscattering intensity vs. wavelength for rat subcutaneous tissue containing a polymer membrane (ΔI(λ)). B) The standard curve used to calculate hemoglobin concentration from the α value, which is a coefficient calculated to relate hemoglobin absorption A(λ) (Figure 3) and (ΔI(λ) through a variation of Beer's law. Different μs values indicate the range of variation in measurements due to differences in the optical scattering properties of the tissue. Figure 12 shows the relative hemoglobin content throughout the tissue containing the polymer membranes. ^Significant difference (p < .05) between the relative hemoglobin concentration for the given membrane and normal subcutaneous tissue, f Significant difference (p < 0.05) between the relative hemoglobin concentration for the given membrane and all other membranes at the given timepoint. PLGA 50:50 membranes were not large enough for 4D-ELF analysis at 10 weeks.

DEFINITIONS

As used herein, the term "subject" refers to any animal (e.g., a mammal), including, but not limited to, humans, non-human primates, rodents, and the like, which is to be the recipient of a particular treatment. Typically, the terms "subject" and "patient" are used interchangeably herein in reference to a human subject.

As used herein, the term "in vitro" refers to an artificial environment and to processes or reactions that occur within an artificial environment. In vitro environments can consist of, but are not limited to, test tubes and cell culture. The term "in vivo" refers to the natural environment (e.g., an animal or a cell) and to processes or reaction that occur within a natural environment. The terms "test compound" and "candidate compound" refer to any chemical entity, pharmaceutical, drug, and the like that is a candidate for use to treat or prevent a disease, illness, sickness, or disorder of bodily function. Test compounds comprise both known and potential therapeutic compounds. A test compound can be determined to be therapeutic by screening using the screening methods of the present invention. In some embodiments of the present invention, test compounds include antisense compounds.

The term "poly(α-hydroxyester)" refers to a collection of biodegradable and bioresorbable polymers, such as polylactides (PLa), and polyglycolides (PGA) which hydrolytically degrade into alpha-hydroxy acids. The present invention is not limited to any particular poly(α-hydroxyester) .

DETAILED DESCRIPTION OF THE INVENTION

Certain illustrative embodiments of the invention are described below. The present invention is not limited to these embodiments. Regenerative medicine and in vivo biosensor applications, for example, require the formation of mature vascular networks for long-term success. In developing embodiments of the present invention, it was investigated whether biodegradable porous membranes induce the formation of a vascularized fibrous capsule and if so, the effect of degradation kinetics on neovascularization. Poly(L-lactic acid) (PLLA) and poly(DL-lactic-co-glycolic) acid (PLGA) membranes were created by a solvent casting/salt leaching method. In developing embodiments of the present invention, PLLA, PLGA 75:25 and PLGA 50:50 polymers were used to vary degradation kinetics. The membranes were designed to have an average 60μm pore diameter.

Membrane samples were imaged by scanning electron microscopy at several time points during in vitro degradation to assess any changes in pore structure. The in vivo performance of the membranes was assessed in Spraque Dawley rats by measuring vascularization within the fibrous capsule that forms adjacent to implants. The vascular density within lOOμm of the membranes was compared to that seen in normal tissue, and to that surrounding the commercially available vascularizing membrane THERACYTE (www.theracyte.com). The hemoglobin content of tissue containing the membranes was measured by 4-dimensional elastic light scattering (4D-ELF) as a novel method to assess tissue perfusion. Results showed that slow degrading membranes induced greater amounts of neovascularization and a thinner fibrous capsule relative to fast degrading membranes. The present invention is not limited to a particular mechanism. Indeed, an understanding of the mechanism is not necessary to practice the present invention. Nonetheless, it is contemplated that these results may be due to both, an initially increased number of macrophages surrounding the slower degrading membranes, and to the maintenance of their initial pore structure.

Membrane characterization

After fabrication, all membranes showed the same type of interconnected pore structure and dimensions (Figure 1). The average pore size of the membranes was 60 ± 20 μm. The average thickness of membranes was 0.71 ± 0.11 mm. At ten weeks of in vitro degradation in PBS, the PLLA membranes maintained their pore structure, while the PLGA 75:25 membranes showed decreases in pore size and changes in surface topography. PLGA 50:50 membranes had become too fragile to handle at this time point so it was not possible to image these samples. However, these membranes began to show changes in surface topography and pore structure at 4 weeks. Regarding membrane degradation rates, the mass loss of PLGA 50:50 was greater than PLGA 75:25 and the mass loss of both these membrane types was greater than that seen in PLLA (Figure 2). The PLLA membranes degraded 7.6 ± 4.1 percent over the ten week time period, while PLGA 75:25 membranes degraded an average of 21 ± 3.9 percent, and PLGA 50:50 membranes degraded 86.5 ± 2.9 percent. At two weeks, the PLLA and PLGA 75:25 samples both showed little to no degradation. PLGA 75:25 only began to show slightly increased degradation over PLLA at the six week time point, with a statistically significant difference occurring at 8 weeks (p< .05).

Characterization of vascularization via histology Sections of subcutaneous tissue obtained with each membrane sample were stained with hematoxylin and eosin (Figure 3). H&E images show that the PLGA 75:25 and PLGA 50:50 membranes experienced similar changes in structure over time. There was evidence of bulk degradation as per the large holes where the polymer was previously present. The fibrous capsule surrounding PLLA membranes was significantly smaller than that surrounding PLGA 50:50 membranes at both time points (Table 1).

Table 1. Fibrous capsule thickness (μm) surrounding membrane implants

The number of blood vessels within lOOμm of the membranes was counted to quantify any differences in vascularization surrounding the different materials. Membranes implanted in the caudal position showed a slight increase in vascular density compared to those implanted in the rostral position, however this difference was not significant (p>.05). Therefore data for all samples of the same material type were grouped and analyzed together for each time point. The vascular density within lOOμm of each polymer membrane type was significantly different (Figure 4). Tissue surrounding PLLA membranes showed the greatest amount of vascularization at both time points, with a vascular density of approximately 750 per mm . Tissue surrounding the fast degrading PLGA 50:50 membranes, on the other hand, showed the least amount of vascularization, with average values ranging between 300-400 vessels per mm . However, even these membranes led to vascular densities similar to those seen in normal subcutaneous tissue. At both time points a clear trend was observed for blood vessel number with respect to degradation rate of the experimental membranes, with the slowest degrading membrane performing the best. The PLLA and PLGA 75:25 membranes also performed significantly better than the THERACYTE control membranes, while the PLGA 50:50 membranes showed a slightly lower, but comparable amount of vascularization to THERACYTE. No significant change was detected in vessel number for samples harvested at two versus ten weeks for all membrane types. However, at ten weeks THERACYTE membranes did not have a significantly higher number of vessels compared to the normal subcutaneous tissue.

Characterization of vascularization via immunohistochemistry

Tissue sections were stained with CD31 and α-smooth muscle actin antibodies in order to differentiate mature vasculature (excluding capillaries) from immature vasculature and capillaries. Immature vessels and capillaries are characterized as those without a pericyte or smooth muscle cell lining, and therefore are stained only brown (CD31), while mature vessels are defined as those having a pericyte or smooth muscle cell lining and are also stained blue (α- SMA). Blood vessel density from membranes is shown in Figure 4. The CD31 labeling also served to verify the data obtained from the H&E staining analysis. The total blood vessel density around each membrane was found to be slightly higher when quantified by immunohistochemsitry compared to the results obtained from H&E staining. However, the differences were not significant and the trends seen with the H&E staining still hold (Figure 5). Although there was little change between the total vascular density at two and ten weeks, the number of mature arteries, arterioles, veins and venules decreases during this timeframe (Figure 7). At two weeks, the number of mature vessels surrounding THERACYTE, PLLA, and PLGA 75:25 membranes was significantly higher (p<.05) than that seen in normal tissue. However, by ten weeks only PLLA and PLGA 75:25 membranes showed significant differences in mature vasculature when compared to the normal tissue. Also, although there was a measurable difference in the total vascular density around the various membranes at ten weeks (Figure 6), the mature vessel density was found to be quite similar for all materials at this time point (Figure 7). Figure 8 shows the contribution of mature vessels and immature vessels and capillaries to the total vascular density for the various membranes. At two weeks a trend is seen in regards to material degradation rate and mature vessel density, with PLLA membranes having a higher percentage of mature vasculature compared to PLGA 75:25 and PLGA 50:50 membranes. At ten weeks, however, the percentage of immature vessels and capillaries increased for PLLA and PLGA 75:25 membranes.

Inflammation surrounding the different membrane materials was also assessed by EDl macrophage staining (Figure 9). At two weeks, the macrophage density surrounding all membrane types was greater than that seen in normal subcutaneous tissue. Results show slight differences in macrophage density between PLLA and PLGA 50:50 at this timepoint. At ten weeks, there was a marked decrease in macrophage number surrounding all membrane materials. At this time point, no difference in macrophage density was found among the different membranes (Figure 10).

Characterization of tissue perfusion via 4D-ELF

The relative hemoglobin content surrounding and within the membranes was measured by 4D-ELF as a method to assess the tissue perfusion supplied by the newly formed vasculature (Figure 12). A typical backscattering spectra obtained from these measurements is shown below (Figure 1 IA). As discussed previously, an α value is calculated by a variation of Beer's law to relate the 4D-ELF backscattering signal and the hemoglobin extinction coefficient. The reported α values are related back to hemoglobin concentration (Figure 1 IB). PLLA, PLGA 75:25, and THERACYTE membranes led to significantly higher hemoglobin levels than the fast degrading PLGA 50:50 membranes at the two week timepoint. PLLA membranes performed better than all other membranes at both timepoints. The trends in the data at the two week timepoint match the trends seen for total blood vessel density surrounding the implants. However, at ten weeks, there is no difference in the hemoglobin concentration surrounding THERACYTE and PLGA membranes.

As previously stated, neovascularization is a crucial process for the success of tissue engineered devices. Research by several groups has shown that porosity and pore size are important factors for controlling vascularization around nondegradable biomaterials (Clowes et al, 1987, Am. J. Surg. 153:501-504; Golden et al, 1990, J. Vase. Surg. 11:838-845; Lam et al, 1995, J. Biomed. Mater. Res. 29:929-942). As such, in developing embodiments for the present invention it was investigated whether changes in the microarchitecture of degradable porous membranes play a role in the neovascularization surrounding such membranes. A degradable vascularizing polymer system would allow for the formation of mature vasculature surrounding an implant, and then degrade such that chronic inflammation would not persist.

It was contemplated that vascularization of the membrane -host tissue interface is modulated by the porous structure of the biodegradable membrane, which changes as a function of membrane degradation kinetics throughout the vessel maturation process. A biodegradable membrane that can maintain the integrity of its 3-dimensional porous structure allows for the greatest amount of neovascularization. Results from an initial in vitro study demonstrated that high molecular weight PLLA membranes were able to maintain their pore structure over the course of the ten week study. The surface area of PLLA membranes was also greater than that of the other membranes at 10 weeks. Contraction and clumping of PLGA 75:25 and 50:50 polymer areas were observed both in vitro and in vivo. PLGA 75:25 membranes began to show changes in pore size and surface topography at 6 weeks in vitro, and PLGA 50:50 membranes showed significant changes in these parameters at four weeks in vitro. Other in vitro studies have also shown that changes in pore size and porosity occur in PLGA membranes over time (Lu et al., 2000, Biomaterials 21:1837-1845). In some embodiments, the average pore size of the biodegradable membranes was 60 μm. However, the scaffold fabrication technique utilized led to a wide pore size distribution and the present invention is not limited to a biodegradable pore size, and more homogeneous membranes are contemplated.

In vivo results demonstrate that the slow degrading PLLA membranes perform significantly better than PLGA 50:50 membranes, and marginally better than PLGA 75:25 membranes in terms of inducing neovascularization within the surrounding fibrous capsule. A significant difference in vascular density was found between PLGA 75:25 membranes and PLGA 50:50 membranes. Thus, at least for the poly(α-hydroxyester) family of polymers, of which all are herein incorporated, there is a correlation between degradation rate and the amount of vascularization surrounding a membrane. A potential explanation for this observation is the difference in the ability of these polymer membranes to maintain their pore structure over time. Another reason for the better performance of slow degrading membranes is that fast degradation causes a more acidic microenvironment, potentially leading to decreased cell viability. Sung et al. have shown that smooth muscle cells seeded on PLGA 50:50 scaffolds have a slightly decreased viability compared to cells seeded on slower degrading ε- polycapro lactone (PCL) (Sung et al., 2004, Biomaterials 25:5735-5742). However, their study focused on the vascularization within a porous scaffold and did not consider the surrounding fibrous capsule. The different chemistries of PCL versus PLGA could have also played a role in their results.

In addition to observing increased neovascularization around slow versus fast degrading polymers, it was found that these biodegradable membranes perform better than, or comparable with, THERACYTE membranes in terms of their ability to promote neovascularization. One potential explanation is that PLLA membranes were constructed with, and maintained, an average 60μm pore size, while the THERACYTE membrane, which is made of PTFE, has 5μm pores. The 60μm pore size has been shown to be optimum in maximizing vascularization in several independent studies (Lam et al., 1995; Sharkawy et al., 1998, J. Biomed. Mater. Res. 40:598-605; Fujimoto et al., 1993, J. App. Biomat. 4:347-354). Also, Lam et al. have shown that vascularization is increased around PLLA scaffolds when compared to PTFE materials with the same pore structure (Lam et al., 1995). Although, the three different layers comprising the THERACYTE membrane make it structurally different from the biodegradable membranes investigated in this study, it served as a "state-of-the-art" reference material. It is noted that the THERACYTE membrane is a commercially available vascularizing membrane with some proprietary specifications. As such, it is demonstrated herein that the use of biodegradable materials as vascularizing membranes for tissue engineering applications where immunoisolation is not a concern.

Another important finding is that there was no significant difference in total vessel density at two and ten weeks. However, staining for α-smooth muscle actin revealed that the number of vessels with a smooth muscle or pericyte lining decreased over the time course of this study. These two findings taken together suggest that the "mature" vessels lost their smooth muscle or pericyte lining over time. The present invention is not limited to a particular mechanism. Indeed, an understanding of the mechanism is not necessary to practice the present invention. Nonetheless, it is contemplated that the process of neovessel regression is taking place at ten weeks, as vessels that are not associated with pericytes have been found to be less stable and more likely to regress. Alternatively, it is also contemplated that new capillaries (which can be stable) develop and maintain perfusion (Fig. 12). It has been shown that growth factor stimulation is not needed to sustain vessels that have a smooth muscle or pericyte lining (Ramsauer and D'Amore, 2002, J. Clin. Inv. 110: 1615-1617). Despite the decrease in mature vasculature, their density surrounding PLLA and PLGA 75:25 membranes was found to be higher than that in normal subcutaneous tissue at this timepoint.

In addition to differences in vascularization around different membrane types, it was found that, on average, the fibrous capsule thickness was smaller for PLLA membranes relative to PLGA 50:50 samples. Regarding potential variability due to implant location, there was generally a slight increase in vascular density, and decrease in capsule thickness in the caudal area relative to the rostral area. This type of difference has also been seen by Picha and Drake (1996, J. Biomed. Mater. Res. 30:305-312), who suggest that because the fat bed is much larger in the caudal position the amount of mechanical stress around the scaffolds is decreased, thus decreasing the amount of collagen production by fibroblasts. The increased neovascularization surrounding porous biodegradable membranes is likely due to the inflammatory response associated with these biomaterials. Several studies have shown that there is a correlation between chronic inflammation and angiogenesis (Jackson et al., 1997, J. Fed. Am. Soc. Exp. Biol. 11:457-465). Also, porous polymer structures show increased levels of inflammation when compared to nonporous constructs (Lam et al., 1995) As macrophages are the primary cell type associated with both inflammation and angiogenesis, we investigated whether there is a difference in macrophage density around different scaffold types. Results from EDl staining for invading macrophages show that at two weeks there are a greater number of macrophages around PLLA membranes. The density of infiltrating macrophages decreases at ten weeks and all material types show similar amounts of macrophages present at this time point. Thus, the severity of the chronic inflammation does subside and equalize over time. Interestingly, there was a correlation between macrophage infiltration and the presence of mature vasculature. The decreased number of macrophages likely corresponds to a decrease in the amount of secreted angiogenic growth factors. It has been shown that decreases in VEGF can lead to loss of the pericyte lining and vessel regression if high levels of Angiotensin 2 are present (Ramsauer and D'Amore, 2002). It has been contemplate that the pore size of the scaffold affects macrophage activation and cytokine secretion. Padera and Colton (1996, Biomaterials 17:277-284) suggest that pore sizes correlating to cellular dimensions allow macrophages to remain in a rounded morphology, which may allow for continued release of angiogenic cytokines. Larger pore sizes, on the other hand, allow cells to spread, which induces a differentiated state and allow for collagen secretion and fibroblast proliferation. In some embodiments, the present invention provides methods for assessing tissue perfusion using 4D-ELF. Other studies have used laser Doppler velocimetry to assess perfusion (Rafael et al, 2000, Cell Trans. 9:107-113; Orlandi et al, 1988, Clin. Sci. 74: 119- 121). However, 4D-ELF is a more promising method, as it also detects levels of oxygenated versus deoxygenated hemoglobin, and gives information about the state of the tissue surrounding the implant. By using different polarizations, the depth at which the tissue is interrogated is also controlled. This parameter is important for assessing the amount of vascularization surrounding an implant. The 4D-ELF method as described herein probed not only the fibrous capsule area, but also some depth into the tissue containing the membrane, including the region within the membrane pores. The results demonstrate that 4D-ELF assesses general microvascular blood content. It should be noted that there are temporal variations in microvascular hematocrit, however the 4D-ELF results correlate with the histological results, suggesting vascular perfusion (Lipowsky, 2005, Microcirc. 12:5-15).

In some embodiments, compositions of the present invention provide slow degrading poly(α-hydroxyl esters) for use as vascularizing membranes in biomedical applications. For example, they allow for increased vascularization and a thinner fibrous capsule when compared to a fast degrading polymer. In further embodiments, the present invention provides for novel methods, such as 4D-ELF, in assessing tissue perfusion in a sample in vitro or in vivo, via hemoglobin content calculations. In some embodiments, the degradable membrane is less than lOOμm in thickness and, for example, serves as the outer layer on cell encapsulation devices or biosensors. In some embodiments, the layer as previously described improves, for example, the functionality and success of biodegradable hollow fibers for vascular networks or guided tissue regeneration.

EXPERIMENTAL

The following examples are provided in order to demonstrate and further illustrate certain preferred embodiments and aspects of the present invention and are not to be construed as limiting the scope thereof.

EXAMPLE 1-Material implantation and observations

Porous poly(α- hydroxyester) membranes were implanted into the subcutaneous tissue of eighteen Spraque Dawley rats. The animals were split equally into three groups based on the type of experimental membranes that they received. The first group received PLLA membranes, the second group PLGA 75:25 membranes, and the third group PLGA 50:50 membranes. Each group received membranes in the right caudal (lower back) and rostral (upper back) positions. All animals had THERACYTE membranes implanted in the left caudal and rostral positions to serve as a benchmark nondegradable vascularizing control material. The amount of vascularization in the fibrous capsule within lOOμm of the perimeter of the implants was quantified by histo morphometry for each group at two and ten weeks. Differences in the vascular density, fibrous capsule thickness, and inflammation surrounding the different membrane types were assessed. Differences in vascularization between the rostral and caudal locations for each material type were also measured to check for variability due to implant location. The hemoglobin content of the harvested tissue was measured by a light scattering spectroscopic technique (4-dimenstional elastic light scattering fingerprinting) to assess tissue perfusion.

EXAMPLE 2-Porous membrane fabrication Membranes of PLLA (MW > 300,000; Polysciences, Inc., Warrington, PA), PLGA

75:25 (MW 90,000-126,000; Sigma, St. Louis, MO), and PLGA 50:50 (inherent viscosity .55- .75; Sigma, St. Louis, MO) were prepared by a salt leaching/solvent casting method. The polymers were first dissolved in dichloromethane under constant stirring. For PLLA, a 5.5% (w/v) solution was used, for PLGA 75:25 a 6.7% (w/v) solution was used, and for PLGA 50:50 a 14.5% (w/v) solution was used. An average 60μm pore size was formed within the membranes by addition of 60-106μm sodium chloride crystals. Salt within this size range was obtained by grinding the crystals with a mortar and pestle and then filtering them between 106 and 60μm sieves. Ninety percent porosity of the membranes was achieved by addition of this percentage of sodium chloride, by weight, to the polymer solutions under constant stirring. The homogenized polymer solutions were cast to an average thickness of 0.7 mm, and placed on dry ice to maintain the membrane structure during solvent evaporation. The membranes were then placed in deionized water to leach out the salt. The water was changed every hour for 6-8 hours. Addition of silver chloride to the water was used to determine whether salt still remained in the membrane. THERACYTE membranes consisted of an immunoisolating membrane, a vascularizing membrane with 5 μm pores, and a woven fiber mesh for support. The total thickness of the THERACYTE is 100 μm. The porosity of the vascularizing membrane is proprietary. Circular disks with a 0.75 cm diameter were punched out from all the porous polymer sheets. Samples were sterilized according to the manufacturer's instructions using an AN74j/Anprolene ethylene oxide sterilization system that performed a 24 hr. degassing step (Anderson Sterilization, Inc. Health Science Park Haw River, NC, USA).

EXAMPLE 3-Membrane characterization

Scanning electron micrographs were taken of the membranes to confirm that an interconnected pore structure with the desired 60μm pore size existed. Cross sections of the polymer membranes were mounted with double sided carbon tape to sample stubs. A sputter coater (Cressington 208HR, England) was used to coat the samples with a 3-5nm thick layer of gold. At least three different samples of each membrane type were viewed and imaged using a Hitachi 3500N Scanning Electron Microscope (Hitachi; Pleasanton, CA). Multiple images were taken of each sample and the average pore diameter from each image was measured using Image Pro Plus 5.0 software.

In vitro degradation of the membranes was studied over a ten week period. Samples were pre-wetted in ethanol and placed in 15 mL of phosphate buffered saline (PBS), pH 7.4. The samples were stored at 37⁰C to simulate physiological conditions. The weights of the membranes were measured on a weekly basis. The samples were rinsed with deionized water to remove residual salts, air dried, and then placed under vacuum overnight. Their dry weight was measured the following day and fresh buffer solution was added to the samples after pre- wetting with ethanol. The buffer solution was not stirred during the week-long incubation. Samples of each membrane type were viewed by scanning electron microscopy on a biweekly basis for any evidence of changes in pore structure over time.

EXAMPLE 4-Animal surgery and tissue harvest

Animals were anesthetized using the inhalant machine Impact 6 (Vetequip Inc., Pleasanton, CA). Isofuorane was given at a concentration of 2% with an oxygen flow rate of 2 L/min. Following anesthesia, the backs of the animals were shaved and the incision sites were disinfected using alcohol and butadiene. An incision, approximately 1.5 cm in length, was made at the implantation site and a subcutaneous pocket created via blunt dissection. The membranes were implanted in the subcutaneous pockets away from the incision. The wounds were closed with surgical staples, which were removed one week post-operation. At two and ten week time points, nine animals were euthanized with CO₂ asphyxiation, followed by bilateral thoracotomy and the samples were harvested. The explants included the tissue that was surrounding the implanted material. Half of each explanted tissue sample was used for 4D-ELF analysis, while the other half was fixed in 10% neutral buffered Accustain formalin solution (Sigma, St. Louis, MO) for 24-48 hours.

EXAMPLE 5- Histological and immunohistochemical characterization of vascularization

The fixed samples were processed through a graded series of 70-100% ethyl alcohol solutions, cleared in xylene, and then embedded in paraffin. Cross sections 5 μm in thickness were cut using a microtome (HM 350 S Microm, Richard-Allan Scientific, Kalamazoo, MI). Prior to embedding, the samples were cut in half, such that cross sections were first obtained from the center of the membranes and then out towards the edge. The sections were placed in a 45⁰C water bath and were floated onto silane coated slides (LabScientific, Livingston, NJ). Four cross sections at random depths within the membranes were obtained from each tissue sample. Two of these cross sections were stained with hematoxylin and eosin (H&E) (Richard Allen Scientific, Kalamazoo, MI).

The reamining two cross sections were used for immunohistochemistry (IHC). IHC was performed using a CD31 (PECAM-I) endothelial cell marker (Abeam, Cambridge, MA) to ensure that total blood vessel counts obtained from H&E stained slides were accurate. In addition, an EDl antibody (Serotec, Raleigh, NC) for infiltrating macrophages was used to detect any chronic inflammation around implants. Finally, a α-smooth muscle actin antibody (Sigma, St. Louis, MO) was used to detect mature arteries, arterioles, veins and venules that are less likely to regress. Capillaries are not included in the mature vessel count, as true capillaries do not express α-smooth muscle actin. Double antibody labeling was performed for CD31 and α-smooth muscle actin markers for an accurate assessment of the percentage of mature vessels, excluding capillaries, to the total vessels within a cross section. For the double labeled sections, a brown horse radish peroxidase (HRP) substrate was used for CD31 and EDl labeling and a blue alkaline phosphatase substrate was used for the α-smooth muscle actin marker.

EXAMPLE 6-Imaging and analysis

Histological slides were viewed using a Nikon TE2000-U inverted light microscope (Nikon USA, Melville, NY). Digital images were obtained with an Olympus Qcolor3 digital camera (Olympus America Inc., Melville, NY) and analyzed with ImagePro Plus 5.0 software (Media Cybernetics, Silver Spring, MD). For each cross section, images were taken in ten regions adjacent to the implanted material along its entire length. For each image, the counting areas were chosen to be within the fibrous capsule area and up to lOOμm from the implant. Three counting areas were randomly chosen along the fibrous capsule length of each image and were digitally marked in Image Pro Plus using an area of lOOμm x lOOμm. Vessels within this area were digitally tagged and results are reported as vessel number per mm² of tissue (mean ± standard deviation).

For H&E images, structures were identified as vessels if they met two of the three following criteria: an endothelial cell lining, a well defined lumen, and the presence of red blood cells. For slides labeled with CD31 and α-smooth muscle actin, structures that stained brown and/or blue and that had a well defined lumen were counted as blood vessels. Vessels with a smooth muscle cell lining (blue stain) were counted as mature, while vessels that only stained brown were considered to be immature vessels or capillaries depending on their size. The percentage of mature vessels within the first lOOμm of the fibrous capsule of each type of implant was quantified. The total number of blood vessels measured by IHC was compared to the numbers obtained from H&E vessel counts as well. For slides labeled with the EDl antibody, the density of macrophages within the first lOOμm of the fibrous capsule of the different membranes was compared. One grader performed all counting analyses.

EXAMPLE 7-Characterization of vascularization using 4-dimensional elastic light scattering spectroscopy (4D-ELF)

Relative hemoglobin levels surrounding and within the implants were assessed by 4D- ELF, a type of light scattering spectroscopy. 4D-ELF is based on the light scattering principle that the intensity of light scattered from a tissue without a change in wavelength is a function of both the light scatterers and absorbers in the tissue. As hemoglobin is the primary molecule that absorbs light in tissue, 4D-ELF can easily distinguish hemoglobin "fingerprints" from those of other molecules. A nominal coefficient, α, is calculated by a least squares regression method to relate A(λ) and the 4D-ELF signal obtained from tissue (ΔI(λ)) through a variation of Beer's law. Specifically, Beer's law is applied with the assumption that the signal attenuation due to optical absorption has an inverse exponential relationship with the absorber (hemoglobin) concentration. This assumption yields the following formula: ΔI(λ) = ΔI_scatteπng(λ) e , where ΔI_scatteπng(λ) represents the spectrum that would be obtained if the sample did not contain any hemoglobin. The ΔI_scatteπng(λ) spectrum is assumed to be a smooth curve, as no signal attenuation would occur without the presence of an absorber. The α coefficient is therefore calculated such that ΔI_scatteπng(λ) does not show any of the characteristic features associated with A(λ). The α value is used as a relative measurement of the hemoglobin concentration. A standard curve of α vs. hemoglobin concentration allows calculation of the actual hemoglobin concentrations in the tissue samples. However, it is more accurate to report the α values themselves (Siegel et al, 2006, App. Optics 45:335-342).

Samples for 4D-ELF analysis were wrapped in saline dampened gauze and stored on ice until measurements were taken (1-2 hours). Three light scattering measurements were taken in different locations for each sample. During a single measurement, a 1mm in diameter circular area of the sample was illuminated by a collimated beam of broadband light emitted from a xenon light source (Oriel Inc., Strafford, CT). The light scattered from the tissue was gathered by the collection system, which consists of an analyzing polarizer that is either co-polarized or cross-polarized with respect to the incident direction of polarization, a Fourier lens (Newport, Inc., Irvine, CA), and an imaging spectrometer coupled to a charge coupled device serving as the detector. The intensity of light scattered for wavelengths between 400-700 nm was recorded at scattering angles within ±2° from the backward direction. A more in depth description of the instrument has been previously published (Roy et al., 2004, Gastroent. 126: 1071-1081).

EXAMPLE 8-Statistical analysis Data analysis was performed using GraphPad Prism 4 (GraphPad Software, SanDiego,

CA). Paired t-tests were performed to detect any differences in vascularization between the rostral and caudal locations. One way ANOVA tests were used to measure differences in vascularization and hemoglobin α values surrounding the different polymer membranes. For ANOVA tests, a post hoc Bonferroni test was performed between groups with significant differences to correct for the multiple pairwise comparisons. A significance level of .05 was used for all tests.

All publications and patents mentioned in the present application are herein incorporated by reference. Various modification and variation of the described methods and compositions of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention that are obvious to those skilled in the relevant fields are intended to be within the scope of the following claims.

Claims

CLAIMSWe claim:

1. A composition comprising a surface erodable biodegradable porous membrane comprising poly (α-hydroxyester) polymers in contact with vascularized tissue.

2. The composition of claim 1 , wherein said poly (α-hydroxyester) polymers comprise poly(L-lactic acid) and poly(DL-lactic-co-glycolic) acid.

3. The composition of claim 1 , wherein said membrane has a pore size of at least 60 microns.

4. The composition of claim 1 , wherein said membrane has a porosity of at least 75 %.

5. A method of inducing vascularization within a tissue comprising implantation of a biodegradable membrane of a surface erodable biodegradable porous membrane comprising poly (α-hydroxyester) polymers in a subject, thereby inducing vascularization within a tissue.

6. The method of claim 5, wherein said poly (α-hydroxyester) polymers comprise poly(L- lactic acid) and poly(DL-lactic-co-glycolic) acid.

7. The method of claim 5, wherein said membrane has a pore size of at least 60 microns.

8. The method of claim 5, wherein said membrane has a porosity of at least

75 %.