Sonographic evaluation of knee cartilage defects implanted with preconditioned scaffolds.

3307jum_online_Layout 1 6/20/14 9:00 AM Page 1241

ORIGINAL RESEARCH

Sonographic Evaluation of Knee Cartilage Defects Implanted With Preconditioned Scaffolds Sleiman R. Ghorayeb, PhD, Adam Levin, MD, Michael Ast, MD, John A. Schwartz, Daniel A. Grande, PhD

Article includes CME test

Objectives—The purpose of this study was to develop a novel method for creating an acellular bioactive scaffold, to prove its efficacy in vivo and in vitro for the augmentation of biological repair, and to confirm that sonographic microscopy is a viable modality for monitoring the healing process of osteochondral defects implanted with preconditioned bioactive scaffolds. Methods—Rabbit marrow stromal cells were retrovirally transduced with either bone morphogenetic protein 7 (BMP-7) or insulinlike growth factor 1 (IGF-1) genes, cultured for 9 weeks in nonwoven poly-L-lactic acid scaffolds, and then frozen and lyophilized. The knees were evaluated at 3, 6, and 12 weeks after surgery using 20-MHz ultrasound and then prepared for routine histologic analysis. B-scans of the extracellular matrix defects were compared to histologic results.

Received June 17, 2013, from the School of Engineering and Applied Sciences, Ultrasound Research Laboratory, Hofstra University, Hempstead, New York USA (S.R.G.); Department of Orthopedic Surgery, North Shore–LIJ Health System, Great Neck, New York USA (A.L., M.A.); and Orthopedics Research Laboratory, Feinstein Institute for Medical Research, North Shore Hospital, Manhasset, New York USA (S.R.G., J.A.S., D.A.G.). Revision requested September 3, 2013. Revised manuscript accepted for publication November 13, 2013. This work was presented in part at the 2010 American Institute of Ultrasound in Medicine Annual Convention; March 24–27; San Diego, California. Address correspondence to Sleiman R. Ghorayeb, PhD, School of Engineering and Applied Sciences, Ultrasound Research Laboratory, Hofstra University, 104 Weed Hall, 133 Hofstra University, Hempstead, NY 11549 USA. E-mail: [email protected], sleiman.r. [email protected] Abbreviations

BMP, bone morphogenetic protein; ECM, extracellular matrix; IGF, insulinlike growth factor; PDGF, platelet-derived growth factor; PLLA, poly-L-lactic acid; ROI, region of interest doi:10.7863/ultra.33.7.1241

Results—Control defects showed a void or a mixture of fibrocartilage tissue. Both types of scaffolds resulted in a higher percentage (both P < .001) of primarily hyaline cartilage tissue with intact articular surfaces. The osteochondral defects were clearly observed in each sonographic signature. There were no differences between images of scaffolds treated with IGF-1 or BMP-7. Extracellular matrix regrowth was found to closely parallel (R2 = 0.968; P < .003) the histologic images. A 3-mm defect depth and a 2.5-mm scaffold thickness were measured on the sonograms, comparing well to actual dimensions. Conclusions—There was a gradual increase in healing bordering the defects for the 3-, 6-, and 12-week samples. Also, we have shown that sonography can aid in monitoring implantation of preconditioned scaffolds in osteochondral defects and thus assessing the healing process and cartilage/bone quality. Key Words—articular cartilage; knee arthrotomy; knee repair; musculoskeletal ultrasound; osteochondral defects; scaffold; sonography of cartilage.

A

rticular cartilage injuries in the knee are painful and disabling, often leading to early degeneration. Their treatment remains a major clinical challenge. Although many attempts have been made to treat these injuries, an ideal solution has yet to be identified. Hyaline cartilage has little intrinsic healing ability and often heals with biomechanically inferior repair tissue.1 Much energy is currently being devoted to the study of methods for enhancing the healing potential of osteochondral defects, as well as determining a reliable method for monitoring the healing response in a noninvasive manner.

©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:1241–1253 | 0278-4297 | www.aium.org


Ghorayeb et al—Assessment of Knee Cartilage Implants With Sonography

Despite multiple biological strategies for repair, microfracture chondroplasty remains the current reference standard.2 Techniques that promote and direct this extrinsic repair mechanism to regenerate hyaline cartilage over fibrocartilage may be translatable to clinical adoption. Recent research has focused on the use of biological scaffolds and tissue engineering to augment the healing response of articular cartilage defects. An ideal scaffold would be biologically compatible, noncytotoxic, biodegradable, and biomechanically stable.3 Such a bioactive scaffold would exploit the extrinsic repair response present in osteochondral defects, be chemotactic to mesenchymal stem cells from the marrow or synovial lining, and direct phenotypic differentiation preferentially toward a cartilage phenotype in the articular regions. For application in clinical practice, such a scaffold must be readily available and reproducible. In addition to the need for improving the ability to heal these disabling lesions, another important boundary has been the ability to noninvasively monitor the healing process. Although several different imaging techniques, such as computed tomography, magnetic resonance imaging, and optical coherence tomography,4 have been used for identifying and observing lesions in the knee, studies have been emerging on the usefulness of sonography in this realm.2,4–12 Sonography can be useful in evaluating the acoustic properties of in vivo human cartilage10,12 and in detecting cartilage degeneration in early osteoarthritis.13–15 We describe a novel method for preparing preconditioned scaffolds using gene therapy to create acellular bioactive scaffolds. These scaffolds would provide migrating host cells with a backbone for attachment and caches of growth factors and morphogens in a physiologic manner. They would provide a permissive environment for cell expansion and guide differentiation. We then tested these scaffolds in a model, using osteochondral defects as a way to evaluate the scaffolds’ in vivo efficacy. Furthermore, we attempted to augment existing work to demonstrate whether sonography can indeed be used to noninvasively monitor the healing process of osteochondral defects implanted with these novel preconditioned bioactive scaffolds. This process was done by measuring the healing (or lack of) of the fibrocartilaginous or hyaline extracellular matrix (ECM) 3, 6, and 12 weeks after nonrepair or a repair procedure to yield a quantitative measure of osteochondral defect healing. Although several studies have examined the relationship between sonography and cartilage biomechanics, to our knowledge, no study to date has evaluated the sequential changes observed on histologic tomography and their correlation with sonographic characteristics.

1242

To assess the ECM state and regrowth within these osteochondral defects, it was necessary first to determine a method of identifying the solid zones within sonograms. In examining the ultrasonic propagation properties in the defects, it has been observed that the latter present themselves in images as either dark amorphous areas or light areas with fibrillarlike inhomogeneous patterns based on their morphologic characteristics, similar to what was depicted by Virén et al,4,11 Chérin et al,6 Gelse et al,7 and Laasanen et al.8 The experiments in those studies were very similar to the work presented here, insofar as the operating frequencies (ranging from 20 to 50 MHz) and sample sizes were concerned, and were comparable in their hypotheses. However, we addressed the evaluation of sonography as a viable imaging technique from a different perspective than those shown in the aforementioned studies. Given that the main issue at hand is to correlate image brightness to the ECM, one should note that the brightness is mainly due to the amount of edema present in the tissue. It is important then to hypothesize this phenomenon. That is, on injury or an induced defect in the knee cartilage, the fluid content in and around the injury site increases, resulting in less scatter per volume and, thus, less brightness. On the other hand, increased protein density becomes prominent with increased healing; thus producing more brightness in the scan. For this reason, a palette was selected that would make the difference between cartilaginous and noncartilaginous zones the clearest. This palette translated the original areas nearing either end of the spectrum to white, representing the densest areas of cartilage, and those in the middle of the original spectrum to black, representing the fluid content and other proteins in the defect structure, with a grayscale gradient separating them depending on the acoustic impedances of those elements. The goal of this study was to measure the healing of the cartilaginous ECM in the 12 weeks following induced knee osteochondral defects to yield a quantitative measure of defect healing. We propose a method that measures image brightness by quantifying the correlation of the latter to the overall content of the ECM within the area where the defect was induced. Results from both ultrasonic interrogation and histologic tomography were then correlated.

Materials and Methods We conducted a 2-part study consisting of in vitro and in vivo components. The initial proof-of-concept testing of the bioactive scaffolds was based on the ability of the scaffolds to alter cell metabolism in a generalized cell model. This aspect was tested in vitro using fibroblasts to determine the scaf-

J Ultrasound Med 2014; 33:1241–1253



folds’ effect on collagen and DNA synthesis, with scaffolds preconditioned with platelet-derived growth factor (PDGF) and insulinlike growth factor 1 (IGF-1). The second part of the study involved sonographic evaluation of these scaffolds in an osteochondral defect rabbit model and examined the differential repair with scaffolds preconditioned with bone morphogenetic protein 7 (BMP-7) or IGF-1. Bioactive Scaffold Testing The LNβ-PDGF-β, LNβ-BMP-7, and LNβ-IGF-1 retroviral vector plasmids were constructed with the retroviral long-terminal repeat driving the expression of the neomycin resistance gene and the rat β-actin enhancer/promoter driving the expression of the PDGF-β and IGF-1 genes. The previously created plasmid LNβ-BMP-7 was digested with HindIII/ClaI, and the BMP-7 coding sequence was replaced with either PDGF-β or IGF-1 to generate the plasmids LNβ-PDGF-β and LNβ-IGF-1. Human umbilical vein endothelial cells were used for creation of complementary DNA constructs for PDGF-β, and human embryonic lung cell lines were the source of complementary DNA for IGF-1. The production of retroviral vector particles was performed according to already established procedures.14 Briefly, LNCX, LNβ-IGF-1, and LNβ-PDGF-β retroviral vector plasmid DNA was used to generate retroviral vector particles using PA317 base-packaging cell lines. PA317 cells were seeded in 9-mL D10 medium in 100-mm dishes at 1 × 106 cells per dish. The cells were transfected with 30 μg of LNβ-IGF-1 or LNβ-PDGF-β using calcium phosphate (Clontech, Palo Alto, CA). After 24 hours, the medium was aspirated off, and the cells were washed with phosphatebuffered saline (pH 7.3) before replacement with 9 mL of fresh D10 medium. Twenty-four hours after washing, the medium was replaced with D10 supplemented with 300μg/mL neomycin analog G418 (Gibco BRL, Grand Island, NY). Stable G418-resistant PA317 cell populations were generated after 7 to 10 days and were used as a source of 24-hour retroviral vector supernatants that were filtered through a 0.22-μm syringe filter (Millipore, Bedford, MA). Four hundred microliters of the supernatant and 1.6 mL of fresh D10 medium supplemented with 8-μg/mL polybrene (Sigma-Aldrich Corp, St Louis, MO) were added to the rat tendon fibroblasts, and the mixture was incubated overnight at 37°C. After the incubation period, the medium was replaced with 2 mL of D10 supplemented with 600μg/mL active gentamicin (Gibco BRL). The rat tendon fibroblasts were isolated from the rotator cuffs of adult Sprague Dawley rats (400–500 g). The fibroblasts were initiated in culture by explant out-


growth. Cells were cultured and expanded in supplemented Dulbecco’s modified Eagle’s medium (Gibco BRL) for 3 passages. After their third passage, the fibroblasts were transduced with either the IGF-1 or PDGF-β gene according to the procedure described by Mason et al.15 Transduction was confirmed with messenger RNA analysis by a reverse transcriptase polymerase chain reaction, using primers specific to IGF-1 and PDGF-β. The cells containing active genes were selected by incorporation of the neomycin resistance gene added to the construct. Growth factor production was confirmed by an enzyme-linked immunosorbent assay. Scaffolds were preconditioned by one of the following cell types: wild-type rat tendon fibroblasts, rat tendon fibroblasts transduced with IGF-1, or rat tendon fibroblasts transduced with PDGF-β. The fibroblasts were seeded onto poly-L-lactic acid (PLLA) scaffolds (2 × 10 × 10 mm) at a concentration of 3.0 × 106 cells per scaffold. The cells were maintained and fed with supplemented Dulbecco’s modified Eagle’s medium for the duration of their growth and kept in a humid 37°C incubator at 5% carbon dioxide. After 3, 6, or 9 weeks, the scaffolds underwent freeze-thaw cycles and remained frozen at –80°C for 6 hours to promote lysis of the seeded cells. After the freeze-thaw cycles, the scaffolds were further lyophilized overnight and stored at –20°C. Preconditioned and nonpreconditioned control PLLA scaffolds were cut into 2.5 × 2.5 × 2-mm squares. Wild-type rat tendon fibroblasts were seeded onto the scaffolds at a concentration of 7.5 × 105 cells per scaffold. Each scaffold was cultured on a 6-well plate in supplemented Dulbecco’s modified Eagle’s medium. The cells were allowed to attach and grow for either 3 or 7 days before measurement of collagen production and DNA synthesis. At each respective time point, the scaffolds were transferred to clean 6-well plates and were pulse labeled with either 25-μL/mL tritiated proline ([3H]proline) for 4 hours or 0.5-μL/mL tritiated thymidine ([3H]thymidine) for 24 hours. The scaffolds were washed 3 times with 2 mL of Hank’s balanced salt solution (Mediatech, Inc, Herndon, VA), followed by papain digestion (Sigma-Aldrich Corp) in a 65°C water bath until all of the samples were completely dissolved. After digestion, 1 mL of cell lysate was transferred into scintillation tubes for counting of either [3H]proline for assessment of collagen synthesis or [3H]thymidine for assessment of DNA synthesis (1900 TR liquid scintillation analyzer; Packard Instruments, Meriden, CT). In the second part of the study, 16 adult male New Zealand White rabbits underwent bilateral knee arthrotomies to create an osteochondral defect (3 mm in diameter × 3.7 mm deep) in the trochlea of each knee. Twenty-four

1243



defects were implanted with either IGF-1- or BMP-7-preconditioned PLLA scaffolds as described. The 8 remaining defects were left unfilled as controls. The knees were then harvested at 3, 6, and 12 weeks after surgery and evaluated both sonographically and histologically. This study was approved by the Institutional Animal Care and Use Committee at the Feinstein Institute for Medical Research. Continuous data were expressed as mean ± SD. Differences in the experimental groups and the control group were calculated with use of 1-, 2-, and 3-way analysis of variance with Tukey post hoc tests (P < .05) by the Biostatistics Unit at the Feinstein Institute for Medical Research. Experimental Sonographic Setup As mentioned above, the second portion of this study was conducted to augment and thus confirm existing studies on whether sonography may be used to noninvasively monitor the healing process of osteochondral defects implanted with the preconditioned bioactive scaffolds. Imaging was performed with a scanning acoustic microscope consisting of an ultrasonic pulser/receiver (model 5800PR; Olympus-Panametrics, Waltham, MA),

an immersion single-element transducer (model V317SM, 20 MHz [actual, 16.5 MHz], 15-MHz, –6-dB bandwidth, 0.25-in diameter, and 1-in spherical focus; Olympus-Panametrics), and a FlexSCAN-C ultrasonic immersion C-scan system (Sonix, Inc, Springfield, VA) with a tank containing phosphate-buffered saline solution as the couplant, which was degassed before measurements, and a computer software system with front wall follower gating capability. The images were created by using successive raw radiofrequency A-scan signals (in a way that mimics a B-scan) acquired at 40 frames per 15 seconds. Given the upper –6-dB frequency of the transducer (24 MHz) and assuming a sound velocity of 1660 m/s in cartilage, the step size of the scanner was set at about 0.0173 mm to conform to the Nyquist rate. Sonographic testing took place first on fresh samples before histologic testing. Each knee ensemble with the joint kept intact was set onto a custom acrylic stand, with the area of interest (the osteochondral defect) placed within the focal point of the transducer. The transducer then scanned across a 12.7-mm (0.5-in) line of interest covering the width of the sample, as shown in Figure 1.16 It is important to note that

Figure 1. Typical rabbit knee articular cartilage defects. A, Immediately after creation. B, Control sample. C, defect (dashed circle) without treatment. D, After treatment with the scaffolds. E and F, Exposed scaffold treatment (dashed circles) showing the superimposed scan line of interest (LOI; red line) used for collection of B-scan images. Reproduced with permission from Jung et al.16

1244




the ultrasound transducer was set up to scan the knees in the axial direction from the inferior to superior aspect of the femur and in a position perpendicular to the knee surface. Also, the reason why such a high nominal operating frequency (20 MHz) was used was 2-fold: (1) the scanning occurred directly on the bone samples with no intermediate layers of soft tissues; and (2) the area of interest was the subsurface superficial content of the samples where the scaffolds were implanted. Since the ultimate goal of this study was to show the feasibility of this technique in noninvasive clinical situations in vivo, in which the measurements are conducted through the skin and other soft tissues, the operating frequency would be lowered to account for propagation in these layers. In such situations, the structures evaluated would still be superficial. Therefore, linear array transducers with frequencies in the range of 7 to 12 MHz, would be the appropriate choice. The attainable high resolution using these types of transducers

allows finer and more detailed anatomic descriptions of the superficial structures. The ability to noninvasively assess the quality and quantity of repair tissue is highly valuable to clinicians for managing patients’ progress before returning to activities of daily living and high-performance sports. Each of the samples was placed in the water tank in the focal plane within the transducer’s cone of insonification. The FlexSCAN-C system was used by setting the gate to examine the subsurface matrix within the line of interest. As an example, Figure 2 shows typical radiofrequency Amode scans (amplitude versus time) from two different knee samples. The gated areas in these A-scans, used for imaging the subsurface matrix, are clearly indicated in these images. The first yellow gate (shown across the first largest peak) is referred to as the “front surface follower” and indicates the top-surface reflection of the sample. The front surface follower will “slave” the next gate position to

Figure 2. Radiofrequency A-scans (amplitude versus time) from two different knee samples showing front surface follower (FSF) and subsurface follower (SSF) gates used to generate the B-scan images of the subsurface scaffold matrix. A, Typical A-scan showing reflections from the front surface (FS) of a treated osteochondral defect with a preconditioned scaffold implant and multiple echo reflections (ER) off the hyaline cartilage tissue generated during repair. B, Typical A-scan showing reflections from the front surface of fibrocartilagenous filling in an untreated osteochondral defect and a back wall reflection (BR) off the end junction of the fibrocartilage fill. A

B


1245



the first crossing of the front surface follower threshold. This slave capability keeps the next gate in the proper location, accounting for uneven or tilted surfaces. The second red gate shown is the “subsurface follower,” which allows tracking of the subsurface matrix, which indicates the presence (or lack of) of a scaffold in the final image. Furthermore, once the gates are set, the system’s software uses a peak detector that determines the maximum signal strength (absolute peak) of the A-mode signal within each gate at each data point. This peak value is represented as intensity in the final B-scan image. The B-scan images were recorded with the FlexScanC software and subsequently compared to histologic sections stained with safranin O and hematoxylin-eosin. However, to maintain a fair comparison between sonographic and histologic images, the latter needed to be converted to gray scale before analyzing the ECM within the defects. This process was done with image analysis software (GNU Image Manipulation Program version GPLv3; www.gimp.org), which simply converts the image to gray scale. Average brightness values were then compared to measurements of the percentage of hyaline or fibrocatilage content in the ECM within a region of interest (ROI) on the sonographic and histologic images, obtained with a developed MATLAB algorithm (The MathWorks, Natick, MA) that implements the Floyd-Steinberg dithering technique.17 This program, in short, converts an image to gray scale and rounds the intensity of each pixel to its nearest extreme, black or white. It then calculates the percentage of white pixels in an ROI with respect to the total number of pixels in the ROI. The size of the ROI did not change much from sample to sample but was randomly selected (and remained almost the same across all samples) to confirm repeatability of the percentage values in the ECM. As a result, the total number of pixels in each set of ROIs varied slightly but was not different from one image to another. On average, the size of the ROI ranged from 310 to 2610 pixels. However, one should note that once the size of the ROI was selected, it was positioned at different locations within the defect, as described in the following section.

3). For samples preconditioned with wild-type rat tendon fibroblasts, only those scaffolds preconditioned for 9 weeks showed increases in [3H]thymidine incorporation over untreated control samples. Experimental scaffolds preconditioned with fibroblasts transduced with either IGF-1 or PBGF-β for 6 or 9 weeks showed increases in [3H]thymidine incorporation over controls at 3 days of culture. No difference was seen after 3 days in culture for any of the experimental groups preconditioned for only 3 weeks compared to each other or untreated controls. Post hoc analysis, however, indicated that scaffolds preconditioned with wild-type fibroblasts for 9 weeks resulted in greater DNA synthesis (P < .05) after 3 days in culture than those preconditioned for 3 or 6 weeks, although no difference was seen between the experimental groups pretreated for 9 weeks. The results after 7 days in culture were largely similar for DNA synthesis to those after 3 days (Figure 4). Scaffolds Figure 3. Synthesis of DNA, as analyzed by incorporation of [3H]thymidine, after 3 days of culture. CPM indicates counts per minute; and WT, wild type. *Significant increase over control (P < .05).

Figure 4. Synthesis of DNA, as analyzed by incorporation of [3H]thymidine, after 7 days of culture. Abbreviations and significance are as in Figure 3.

Results Bioactive Scaffold Results Synthesis of DNA, as analyzed by incorporation of [3H]thymidine, showed increases (both P < .001) in the preconditioned groups over control scaffolds not preconditioned after both 3 and 7 days of culture. At the 3-day time point, the increases in DNA synthesis were largely seen in the scaffolds pretreated for longer durations (Figure

1246




preconditioned with wild-type rat tendon fibroblasts for 3, 6, and 9 weeks showed increases (P < .05) in [3H]thymidine incorporation compared to untreated control scaffolds. For samples preconditioned with fibroblasts transduced with IGF-1 for 6 or 9 weeks, an increase (P < .05) in [3H]thymidine incorporation was shown over control scaffolds at the 7-day time point. The experimental group preconditioned with fibroblasts transduced with PDGF-β showed increases (P < .05) in DNA synthesis over untreated controls at 7 days in culture for only those preconditioned for 9 weeks. At this time point, no differences were shown between experimental groups preconditioned for 9 weeks. After both 3 and 7 days of culture, preconditioned samples of all experimental groups showed increased (both P < .0001) [3H]proline incorporation over untreated control scaffolds. By 3 days of culture, scaffolds preconditioned with wild-type rat tendon fibroblasts for 3 or 9 weeks Figure 5. Synthesis of collagen, as analyzed by incorporation of [3H]proline, after 3 days of culture. Abbreviations and significance are as in Figure 3.

Figure 6. Synthesis of collagen, as analyzed by incorporation of [3H]proline, after 7 days of culture. Abbreviations and significance are as in Figure 3.

showed greater (P < .05) collagen synthesis than control samples (Figure 5). Although increases over controls were not significant at this time point for scaffolds preconditioned with wild-type fibroblasts for 6 weeks, there were no differences in [3H]proline incorporation after 3 days of culture between the different pretreatment durations. After 3 days of culture, scaffolds preconditioned with fibroblasts transduced with IGF-1 showed greater (P < .05) collagen synthesis than untreated scaffolds, which was independent of the pretreatment duration. However, only samples preconditioned with fibroblasts transduced with PDGF-β for 6 weeks showed increases in [3H]proline incorporation over scaffolds not preconditioned. Interestingly, there were no differences found between collagen synthesis at 3 days of culture and those preconditioned for 6 and 9 weeks. By 7 days in culture, preconditioned scaffolds in all treatment groups showed increases (P < .05) in [3H]proline incorporation over those of the control scaffolds (Figure 6). These increases were found for all durations of pretreatment studied. At this time point, no changes in collagen synthesis were found with an increasing preconditioning duration. However, in samples preconditioned for 9 weeks, post hoc analysis indicated that scaffolds preconditioned with rat tendon fibroblasts transduced with either IGF-1 or PDGF-β showed greater (P < .05) [3H]proline incorporation than those involving wild-type fibroblasts after 7 days in culture. Both types of preconditioned scaffolds resulted in a higher percentage of primarily hyaline cartilage tissue with intact articular surfaces. Each of the 3 groups of osteochondral defects was evaluated macroscopically, sonographically, and histologically. The control untreated defects were characterized by either an empty edemic cavity (Figure 7B) or a quasi-homogeneous looking mixture of fibrocartilaginous and fibrous tissue (Figure 8B). The preconditioned scaffolds were able to produce more normalappearing cartilage layers (Figure 9, B and D), without the addition of autogenous or allogeneic chondrocytes. Additionally, the regeneration of a quasi-homogeneous subchondral plate and bone was enhanced in these groups at the 12th week compared to controls, as shown in Figure 9F. Sonographic Results The site of the osteochondral defect was clearly identified in each of the quasi-B-scans performed. An unfilled defect that remained devoid of repair tissue 3 weeks later showed an absence of cartilage filling on the sonographic evaluation (Figure 7A), indicating the presence of fluids in the defect structure. By 6 weeks after surgery, the empty osteochondral lesion was filled with fibrocartilaginous repair tissue (arrow),


1247



as seen in histologic studies (Figure 8B). This appearance correlated with an identifiable ultrasonic reflection (arrow) on the front surface of the fibrocartilaginous repair tissue (Figure 8A) but showed more of a hypoechoic behavior in association with the quasi-homogeneous nature of the fibrocartilagenous repair tissue. Sonograms containing defects treated with a preconditioned scaffold implant displayed filling in the defect area within the groove of the femur (Figure 9, A and C). The lack of hyperechoic reflections due to empty defects was distinctly visible in the control images (Figures 7A and 8A) compared to those containing the scaffold-filled defects 3 or 6 weeks into the repair process (Figure 9, B and D). However, at 12 weeks, the BMP-7-filled defect showed well-defined quasihomogeneous hyaline repair (Figure 9F), causing the corresponding sonogram to again show hypoechoic features similar to those shown in Figure 8A, indicative of a quasihomogeneous pattern. Despite evidence of progressive remodeling of the repair tissue with increasing postsurgical time, there were no physical differences evident between sonograms of a scaffold preconditioned with BMP-7 or IGF-1 gene-enhanced cells at 3 versus 6 weeks after surgery. It is important to note, however, that ultrasonic interrogation of all 24 samples containing the scaffold-filled defects showed similar outcomes. To our knowledge, there are no references that discuss ultrasonic interrogation of knee cartilage repair using implanted preconditioned gene-enhanced scaffolds. What we are demonstrating in this study is that

ultrasound does not discriminate between the types of gene therapies such as BMP-7 and IGF-1. Reflections emanating from the scaffold-filled defects showed similar hyperechoic features and similar outcomes. The only difference was observed when either complete or partial hyaline cartilage regeneration took place, which was indicated by a quasi-homogeneous trend, as opposed to inhomogeneous and anisotropic features, as those shown by the scaffolds. We would expect that both molecules (BMP7 and IGF-1) would have a similar regenerative capacity in this application.14,15 The images shown in this study correspond to cropped sections of slightly larger collected images to focus only on the region where the defects took place. The scale bar (ImageJ version 1.45s; National Institutes of Health, Bethesda, MD) shown applies to both the horizontal (scan direction) and vertical (depth direction) axes. The ECM within the defects was analyzed by using an ROI box similar to the one shown in Figure 7. Thus, the acoustic impedances, which depend on the material density and velocity of sound of the features in the ROI, greatly depended on the content of the defects, which varied from sample to sample and when compared to the control knees. The exact locations of these ROIs varied with each sample but were similar in each as far as proximity within the defect site. It is important to note, however, that several ROIs were placed at various locations to confirm the position accuracy and therefore the assessment of ECM content in the appropriate region.

Figure 7. A, Sonogram showing an unfilled cavity (arrow) and reflections off the lateral and medial femoral condyles (LFC and MFC). B, Safranin Ostained section of an untreated osteochondral defect that remained unfilled after 3 weeks (arrow). Also shown in both images is a representative ROI used to measure the ECM percentage.

1248




As a confirmation of the hypothesis stated earlier, the degree of hypoechogenicity decreased with repair time, indicating an increase in the proliferation of cartilage and other proteins within the ECMs of the knee specimens. The percentage of ECM in the self-repaired control sample 6 weeks after defect induction was approximately 86% greater compared to the original control matrix composition, as shown in Figure 10A. The specimens filled with the preconditioned scaffolds showed a progressive regrowth of ECM density during the 3-, 6-, and 12-week healing periods, with approximate average cartilage and hyaline percent compositions of 39%, 47%, and 93%, respectively, which constitute a substantial increase over the control average matrix content. A similar trend can be observed in the histologic counterparts (Figure 10B). The scaffoldfilled samples had average regrowth ECM percentages of 46%, 53%, and 86%, for the 3-, 6-, and 12-week periods. The naturally occurring fibrocartilagenous repair in the 6-week control sample showed an increase of almost 98% improvement over the empty control. Finally, to determine the relationship between the ECM concentrations in the sonograms and those found in the histologic images, regression analysis was used. Coefficient of determination (R2) and P values for the significance of the regression models were calculated. Extracellular matrix concentrations in both modalities were found to be statistically significant (P < .003). Plotting the two factors with respect to each other, rather than repair

time, yielded a linear regression with an R2 value of 0.968, indicating a significant correlation, as shown in Figure 11.

Discussion Although tissue engineering has provided an exciting new pathway in the treatment of articular cartilage lesions, surgeons and scientists still struggle with improving the efficacy and clinical compatibility of this emerging technology. The challenge remains to develop a convenient, clinically applicable, and therapeutically effective method to augment the body’s healing process, without subjecting the patient to unnecessary risk or prohibitive costs. Also, the ability to safely and effectively monitor the response to such treatments remains an area requiring further investigation. Our goals were to validate the use of a novel acellular, preconditioned scaffold in an in vitro model and evaluate its potential for use in an in vivo osteochondral defect model. Furthermore, we aimed to demonstrate whether sonography could be used as a noninvasive method to monitor and assess osteochondral defect repair tissue. We note limitations of this study. The first was our application of the scaffold, which was validated by using rat tendon fibroblasts in an in vitro model for use in an in vivo osteochondral defect model. Although tendon and cartilage are different tissue types with different inherent healing properties, both are derived from a mesenchymal lineage18,19 and respond to similar growth factors.13,14,20

Figure 8. A, Sonogram of an untreated osteochondral defect showing a front surface echo (arrow) of fibrocartilagenous filling after 6 weeks and reflections (yellow circle) off the back wall of the fill. Reflections off the lateral and medial femoral condyles (LFC and MFC) are also indicated. B, Corresponding Safranin O-stained section of the untreated osteochondral defect showing the fibrocartilaginous filling (arrow).


1249



Also, it has been noted that tendon and bone healing, as well as osteochondral defect repair tissue, often shows a fibrocartilaginous tissue type during certain phases of intrinsic repair.19,21,22 Furthermore, several studies have suggested that repair of each of these tissue types may be enhanced

by augmentation of their biological environment.14,23 Second, our in vivo application of the scaffold used BMP-7, as opposed to the PDGF-β used in our in vitro validation. This choice was made because, in the setting of an osteochondral defect, BMP-7 is efficacious in the augmentation

Figure 9. A, C, and E, Sonograms of preconditioned scaffolds filling osteochondral defects (arrows): A, treated with BMP-7 genes after 3 weeks; C, treated with IGF-1 genes after 6 weeks; E, treated with BMP-7 genes after 12 weeks. Echoes show no differences with treatment type. Reflections off the lateral and medial femoral condyles (LFC and MFC) are also clearly shown. B, D, and F, Corresponding Safranin O-stained sections: B, BMP-7 treatment after 3 weeks; D, IGF-1 treatment after 6 weeks; F, BMP-7 treatment after 12 weeks. Continued remodeling and changes in surface morphologic characteristics (arrows) are shown. Note the sonogram in Eshowing front surface echoes (arrows) of partial hyaline cartilage reconstruction filling after 12 weeks and reflections (yellow circle) off the border of the incomplete reconstructed area shown in F (continued).

1250




of repair.14 Finally, we still hope to study how our in vivo data would be affected by the addition of a control-type scaffold, such as a scaffold not preconditioned or one preconditioned with wild-type cells alone. We intended to use many of these principles to enhance the quality of tissue-engineered scaffolds in the hopes of developing an “off-the-shelf” product for the augmentation of biological repair. The PLLA scaffolds were preconditioned with rat tendon fibroblasts for varying times to allow the cells to produce and lay down matrix products into the scaffold. After lysing the cells and lyophilizing the scaffold, the scaffolds were again seeded with fibroblasts and analyzed for DNA synthesis and collagen production to assess whether the preconditioned scaffolds afforded more favorable conditions for growth and matrix production. Furthermore, the study examined the effects of gene therapy on preconditioning these scaffolds by using cells that were gene enhanced with either IGF-1 or PDGF-β for the initial pretreatment. Our results demonstrated that the cells grown on preconditioned scaffolds had greater DNA and collagen synthesis than scaffolds that were not preconditioned. This novel preconditioned scaffold enhanced the repair of an osteochondral defect, with histologically superior repair tissue, when compared to no treatment. Previous studies have suggested that the spontaneous repair of unfilled osteochondral defects resulted in fibrocartilaginous tissue.19,24 Our results are consistent with prior reports in documenting an improvement in the qual-


ity of repair tissue after augmentation of the biological environment.25 Our analysis also demonstrated the ability of sonography to be used as a reliable and reproducible method for evaluating the healing response of these osteochondral defects, which correlated well with our histologic observations in a noninvasive manner. Sequential sonographic B-scan slices were recorded, and gray scale brightness in selected ROIs in the ECM was used to compare their content against histologic examinations performed on the same samples. We show, for the first time to our knowledge, a high correlation between the percentage of the ECM concentration resulting from brightness distributions on both sonographic and histologic images. A very interesting feature that confirms the viability of sonography can be seen when comparing Figures 8 and 9. The fibrocartilagenous tissue that filled the osteochondral defect (Figure 8B) and the regeneration of the quasihomogeneous subchondral plate and bone (Figure 9F) were translated into sonograms containing only front surface reflections (Figures 8A and 9E, arrows) and noticeable hypoechoic features below them. Succeeding major reflections can then be seen at the back wall junction of the fibrocartilage fill (Figure 8, yellow circles) and the area bordering the regenerated hyaline and partially reconstructed inhomogeneous region (Figure 9, E and F, yellow circles). This appearance is typical of quasi-homogeneous/ quasi-isotropic media, which is the case in this fibrocartilagenous tissue (Figure 8B) and hyaline cartilage (Figure 9F).

1251



In this situation, a characteristic A-scan would be similar to that shown in Figure 2B, where only two major reflections (front surface [FS] and back wall reflection [BR]) are present. However, the scaffold-filled defects (Figure 9, B and D) showed more of a hyperechoic matrix due to the anisotropic structure of the scaffold. As can be seen in Figure 10. Average percentages of ECM content for the 3-, 6-, and 12week periods after preconditioned BMP-7 and IGF-1 scaffold repair and for a 6-week self-repair of a control sample, obtained by processing sonograms with the MATLAB algorithm. A, Sonographic results. B, Histologic results. A

B

Figure 11. Regression analysis relationship between the ECM concentrations on the sonograms and those found on the histologic images.

1252

Figure 9, A and C, the content of the osteochondral defects is displayed as a wavy structural organization, indicating the presence of the hyaline cartilage tissue generated as a result of the repair that took place. Here, a typical A-scan would be similar to that shown in Figure 2A, where again a front surface reflection is present but now is indicative of the front wall of the repaired cartilage and subsequent multiple echo reflections stemming from the hyaline cartilage tissue generated during repair. We acknowledge further limitations of this study. Although the results of the in vivo study are encouraging, and the proposed scaffolds seem to have some potential, there are several issues of the in vitro sonographic portion that need to be clarified, specifically as they pertain to the method used as well as to the quantification of the results obtained. Since the aim of the study was to evaluate whether sonography may be used noninvasively to evaluate the integrity of the cartilage lesion, one might question the motivation behind why this study was not conducted using clinically available equipment. The reason was 2-fold: (1) since we were dealing with exposed knees that underwent bilateral arthrotomies, it was technically easier to couple the sonographic field to the uneven sample surfaces by using a scanning acoustic microscope tank filled with saline as opposed to a clinical device in which gel is required; and (2) since histologic examinations were performed on the samples after the sonographic examinations, it was easier to visually correlate the images obtained with the latter to histologic sections. As far as quantification of the ultrasonic signal is concerned, when B-mode images are analyzed and compared with histologic sections, it is common practice to calculate parameters such as the reflection coefficient, integrated reflection coefficient, apparent integrated backscattering, and ultrasound roughness index, so that a comparison can be established with previous studies of sonographic evaluation of articular cartilage and subchondral bone.6,7,9,10 On the basis of what was discussed earlier, our objective was to address the evaluation of sonography as a viable imaging technique from a different perspective than those shown in the aforementioned studies and to demonstrate this proof-of-concept sonographic modality as a potential for clinical applications. It is our hope to provide additional conclusive parameters from the study to be justified in future work. Furthermore, the lack of adequate quantitative reference techniques for evaluation of the compositional and structural integrity and mechanical properties of the osteochondral repairs further limits the comparison of these findings to other studies in which biomechanical testing of the repaired articular cartilage was performed before sonographic analysis. It is our intention




in subsequent studies to assess the overall ECM content of the repaired cartilage and compare it to adjacent intact cartilage as well as to biomechanical properties. Hopefully, these quantitative measures of cartilage healing using the ECM content of the injury site will further confirm the viability of ultrasonic interrogation of spontaneously repaired osteochondral defects. In conclusion, there was a gradual but progressive increase in healing bordering the defects for the 3-, 6-, and 12-week samples. We were able to demonstrate the in vivo and in vitro efficacy of a novel method for preparing preconditioned scaffolds in improving the quality of repair tissue in an osteochondral defect model. Also, we have shown that sonographic results correlate well with corresponding histologic sections and confirm that sonography can aid in monitoring implantation of preconditioned scaffolds in osteochondral defects and thus assessing the healing process and cartilage/bone quality.

References 1.

Khan IM, Gilbert SJ, Singhrao SK, Duance VC, Archer CW. Cartilage integration: evaluation of the reasons for failure of integration during cartilage repair. A review. Eur Cell Mater 2008; 16:26–39. 2. Koplas M, Schils J, Sundaram M. The painful knee: choosing the right imaging test. Cleve Clin J Med 2008; 75:377–384. 3. Safran MR, Kim H, Zaffagnini S. The use of scaffolds in the management of articular cartilage injury. J Am Acad Orthop Surg 2008; 16:306–311. 4. Virén T, Huang YP, Saarakkala S, et al. Comparison of ultrasound and optical coherence tomography techniques for evaluation of integrity of spontaneously repaired horse cartilage. J Med Eng Technol 2012; 36:185– 192. 5. Chérin E, Saïed A, Laugier P, Netter P, Berger G. Evaluation of acoustical parameter sensitivity to age-related and osteoarthritic changes in articular cartilage using 50-MHz ultrasound. Ultrasound Med Biol 1998; 24:341– 354. 6. Chérin E, Saïed A, Pellaumail B, et al. Assessment of rat articular cartilage maturation using 50-MHz quantitative ultrasonography. Osteoarthritis Cartilage 2001; 9:178–186. 7. Gelse K, Olk A, Eichhorn S, Swoboda B, Schoene M, Raum K. Quantitative ultrasound biomicroscopy for the analysis of healthy and repair cartilage tissue. Eur Cell Mater 2010; 19:58–71. 8. Laasanen MS, Töyräs J, Vasara A, et al. Quantitative ultrasound imaging of spontaneous repair of porcine cartilage. Osteoarthritis Cartilage 2006; 14:258–263. 9. Saarakkala S, Toyras J, Hirvonen J, Laasanen MS, Lappalainen R, Jurvelin JS. Ultrasonic quantitation of superficial degradation of articular cartilage. Ultrasound Med Biol 2004; 30:783–792. 10. Saarakkala S, Laasanen MS, Jurvelin JS, Töyräs J. Quantitative ultrasound imaging detects degenerative changes in articular cartilage surface and subchondral bone. Phys Med Biol 2006; 51:5333–5346. J Ultrasound Med 2014; 33:1241–1253

11. Virén T, Saarakkala S, Jurvelin JS, et al. Quantitative evaluation of spontaneously and surgically repaired rabbit articular cartilage using intraarticular ultrasound method in situ. Ultrasound Med Biol 2010; 36: 833–839. 12. Hattori K, Takakura Y, Tanaka Y, et al. Quantitative ultrasound can assess living human cartilage. J Bone Joint Surg Am 2006; 88:201–212. 13. Rodeo SA, Potter HG, Kawamura S, Turner AS, Kim HJ, Atkinson BL. Biologic augmentation of rotator cuff tendon-healing with use of a mixture of osteoinductive growth factors. J Bone Joint Surg Am 2007; 89:2485– 2497. 14. Grande DA, Mason J, Light E, Dines JS. Stem cells as platforms for delivery of genes to enhance cartilage repair. J Bone Joint Surg Am2003; 85:111– 116. 15. Mason JM, Breitbart AS, Barcia M, Porti D, Pergolizzi RG, Grande DA. Cartilage and bone regeneration using gene-enhanced tissue engineering. Clin Orthop Relat Res 2000; 379(suppl):S171–S178. 16. Jung Y, Park MS, Lee JW, Kim YH, Kim SH, Kim SH. Cartilage regeneration with highly-elastic three-dimensional scaffolds prepared from biodegradable poly(L-lactide-co-ε-caprolactone). Biomaterials 2008; 25:4630–4636. 17. Bankman I. Handbook of Medical Imaging: Processing and Analysis Management. New York, NY: Academic Press, 2000. 18. Lietman SA, Miyamoto S, Brown PR, Inoue N, Reddi AH. The temporal sequence of spontaneous repair of osteochondral defects in the knees of rabbits is dependent on the geometry of the defect. J Bone Joint Surg Br 2002; 84:600–606. 19. Shapiro F, Koide S, Glimcher MJ. Cell origin and differentiation in the repair of full-thickness defects of articular cartilage. J Bone Joint Surg Am 1993; 75:532–553. 20. Dines JS, Grande DA, Dines DM. Tissue engineering and rotator cuff tendon healing. J Shoulder Elbow Surg 2007; 16(suppl):S204–S207. 21. Gulotta LV, Rodeo SA. Growth factors for rotator cuff repair. Clin Sports Med 2009; 28:13–23. 22. Hunziker EB. Articular cartilage repair: basic science and clinical progress—a review of the current status and prospects. Osteoarthritis Cartilage 2002; 10:432–463. 23. Kovacevic D, Rodeo SA. Biological augmentation of rotator cuff tendon repair. Clin Orthop Relat Res 2008; 466:622–633. 24. Steadman JR, Rodkey WG, Briggs KK. Microfracture to treat fullthickness chondral defects. J Knee Surg 2002; 3:170–176. 25. Kessler MW, Grande DA. Tissue engineering and cartilage. Organogenesis 2008; 4:28–32.

1253

MRI EVALUATION OF KNEE CARTILAGE.

CORRELATION BETWEEN MICRO-CT SECTIONS AND HISTOLOGICAL SECTIONS OF MOUSE SKULL DEFECTS IMPLANTED WITH ENGINEERED CARTILAGE.

Particulated articular cartilage for symptomatic chondral defects of the knee.

Regeneration of rabbit calvarial defects using cells-implanted nano-hydroxyapatite coated silk scaffolds.

Computational modelling of ovine critical-sized tibial defects with implanted scaffolds and prediction of the safety of fixator removal.

Sonographic Examination of Knee Ligaments.

Quadriceps femoris strength and sagittal-plane knee biomechanics during stair ascent in individuals with articular cartilage defects in the knee.

Nanotechnology biomimetic cartilage regenerative scaffolds.

Cartilage Injuries in the Adult Knee: Evaluation and Management.

An "in vitro" experimental model to predict the mechanical behavior of macroporous scaffolds implanted in articular cartilage.

Fresh Osteochondral Allograft Transplantation for Treatment of Articular Cartilage Defects of the Knee.

Treatment of cartilage defects of the knee: expanding on the existing algorithm.

The surgical management of symptomatic articular cartilage defects of the knee: Consensus statements from United Kingdom knee surgeons.

Articular cartilage defects of the knee: correlation between magnetic resonance imaging and gross pathology.

Microfracture for the treatment of cartilage defects in the knee joint - A golden standard?

Revision cartilage cell transplantation for failed autologous chondrocyte transplantation in chronic osteochondral defects of the knee.

Sonographic evaluation of children with congenital hypothyroidism.

Sonographic evaluation of femoral condylar cartilage in osteoarthritis and rheumatoid arthritis.

Electrospun cartilage-derived matrix scaffolds for cartilage tissue engineering.

[Focal cartilage defects within the medial knee compartment. predictors for osteoarthritis progression].

Acetabular cartilage defects cause altered hip and knee joint coordination variability during gait.

Bioactive Microsphere-Based Scaffolds Containing Decellularized Cartilage.

Repair of retropatellar cartilage defects in the knee with microfracture and a cell-free polymer-based implant.

ECM scaffolds for cartilage regeneration.