UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2021
Cationic contrast-enhanced computed tomography distinguishes between
reparative, degenerative and healthy equine articular cartilage
Nelson, Brad B
Wiley
Tieteelliset aikakauslehtiartikkelit
© 2020 Orthopaedic Research Society All rights reserved
http://dx.doi.org/10.1002/jor.24894
https://erepo.uef.fi/handle/123456789/24418
Downloaded from University of Eastern Finland's eRepository
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/jor.24894.
Acc ep ted Art icle
Brad Nelson ORCID iD: 0000-0002-0205-418X Janne Mäkelä ORCID iD: 0000-0002-6123-1262 Amit Patwa ORCID iD: 0000-0002-9323-3262 Mark Grinstaff ORCID iD: 0000-0002-5453-3668
Cationic contrast-enhanced computed tomography
distinguishes between reparative, degenerative and healthy equine articular cartilage
Brad B. Nelson1, Janne T. A. Mäkelä2,3,4, Taylor B. Lawson2,4, Amit N. Patwa4,5, Brian D. Snyder2, C. Wayne McIlwraith1, Mark W. Grinstaff4, Laurie R. Goodrich1, and Chris E. Kawcak1*
1Orthopaedic Research Center, C. Wayne McIlwraith Translational Medicine Institute, Colorado State University, Fort Collins, CO, 2Center for Advanced Orthopaedic Studies, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, 3Department of Applied Physics, University of Eastern Finland, Kuopio, Finland, 4Departments of Chemistry and Biomedical Engineering, Boston
Acc ep ted Art icle
University, Boston, MA, 5Deparment of Chemistry, School of Science, Navrachana University, Vadodara 391410, Gujarat, India.
*Corresponding Author:
Christopher Kawcak, DVM, PhD, DACVS, DACVSMR
300 W. Drake Road
Fort Collins, CO 80523
Phone:
Email: Christopher.Kawcak@colostate.edu
Running title: Cationic CT of Equine Cartilage
Author contributions statement: BBN, LRG and CEK designed the study. Data was acquired by BBN, JTAM, TBL and ANP. BN analyzed the results and drafted the manuscript. Data was interpreted by BBN, JTAM, BDS, CWM, MWG, LRG and CEK. All authors critically revised the manuscript and approved the final version.
Acc ep ted Art icle
Abstract
Cationic contrast-enhanced computed tomography (CECT) is a quantitative imaging technique that characterizes articular cartilage, though its efficacy in differentiating repair tissue from other disease states is undetermined. We hypothesized that cationic CECT attenuation will distinguish between reparative, degenerative and healthy equine articular cartilage and will reflect biochemical, mechanical, and histologic properties. Chondral defects were created in vivo on equine
femoropatellar joint surfaces. Within defects, calcified cartilage was retained (repair 1) or removed (repair 2). At sacrifice, plugs were collected from within defects, and at locations bordering (adjacent site) and remote to defects along with site-matched controls. Articular cartilage was analyzed via CECT using CA4+ to assess
glycosaminoglycan (GAG) content, compressive modulus (Eeq) and International Cartilage Repair Society (ICRS) II histologic score. Comparisons of variables were made between sites using mixed model analysis and between variables with correlations. Cationic CECT attenuation was significantly lower in repair 1 (1478 ± 333 Hounsfield units [HUs]), repair 2 (1229 ± 191 HUs) and adjacent (2139 ± 336 HUs) sites when compared to site-matched controls (2587 ± 298, 2505 ± 184 and 2563 ± 538 HUs, respectively; all P<0.0001). Cationic CECT attenuation was
significantly higher at remote sites (2928 ± 420 HUs) compared to repair 1, repair 2 and adjacent sites (all P<0.0001). Cationic CECT attenuation correlated with ICRS II score (r = 0.79), GAG (r = 0.76) and Eeq (r = 0.71) (all P<0.0001). Cationic CECT
Acc ep ted Art icle
distinguishes between reparative, degenerative and healthy articular cartilage and highly correlates with biochemical, mechanical and histological tissue properties.
Key words: osteoarthritis, imaging, CT, biomarker, glycosaminoglycan
Introduction
Articular cartilage degeneration and the development of osteoarthritis are substantial concerns in humans as well as horses;1-4 however early detection of injury and monitoring of healing remain diagnostic challenges. One of the sentinel changes in articular cartilage characteristic of osteoarthritis is the loss of
glycosaminoglycan (GAG) content within the extracellular matrix.2,5 Alongside GAG depletion and disruption of the collagen network, articular cartilage becomes structurally weaker with an inability to sufficiently resist loading forces. Coupled with an incapability to self-regenerate, the degeneration of articular cartilage perpetuates inflammation and joint degradation.3 Therefore, early detection of articular cartilage damage prior to advanced joint degeneration is essential for successful outcomes.
Currently, magnetic resonance imaging (MRI) is the optimal method for the evaluation and monitoring of articular cartilage health. While MRI provides
Acc ep ted Art icle
volumetric and whole joint assessment without ionizing radiation, its high cost prevents widespread use in initial clinical assessments or during subsequent follow- up examinations to characterize the progression of healing.4 Some specialized quantitative MRI techniques (delayed gadolinium enhanced MRI of cartilage and T2 mapping) are used to more thoroughly evaluate articular cartilage;4,7,8 however, logistical challenges and computational expertise usually limit their integration in routine clinical practice. Therefore, alternative imaging techniques are warranted to investigate new methods that detect early articular cartilage injury and are capable of monitoring healing.7
Iodinated contrast agents are used to image articular cartilage with computed tomography a technique referred to as contrast-enhanced computed tomography (CECT).6,9 The amount of contrast agent that enters the extracellular matrix depends on the fixed negative charge from GAGs.5,10 Conventional iodinated agents are negatively (or neutrally) charged and are repelled. Cationic agents are positively charged and are electrostatically attracted to the negatively charged GAGs allowing for a higher affinity for articular cartilage at lower iodine doses.
Correlations of cationic CECT attenuation and GAG concentration are strong (r = 0.79 to 0.93)11-15 and surpass conventional anionic agents (r = -0.44 to -0.79).11-
13,16,17 Despite the potential benefit, the ability for cationic CECT to characterize reparative tissue is unknown and these data would be valuable to verify the
integrity of articular cartilage health. Equine translational research models provide
Acc ep ted Art icle
a spectra of cartilage disease states and the ability to compare these data to site- matched controls enables a robust generation of healthy degenerative and
reparative cartilage tissue for ex vivo analysis.18-20 The objective of this study is to examine the capacity of cationic CECT to distinguish between a continuum of
articular cartilage disease states. We hypothesize that cationic CECT will distinguish between reparative, degenerative and healthy equine articular cartilage and is representative of biochemical, mechanical, and histologic tissue properties.
Methods Study design
All in vivo experimental protocols were approved by the Animal Care and Use Committee at Colorado State University (Protocol ID: 14-4907A). Seven juvenile and skeletally mature mixed breed horses aged 2-5 years (6 females, 1 castrated male) weighing mean ± standard deviation (s.d.) of 388.9 ± 49.6 kg were included.
Inclusion criteria required there was no lameness, femoropatellar joint effusion or radiographic evidence of joint disease.21,22
General anesthesia was induced with ketamine (2.2 mg/kg IV) and diazepam (0.1 mg/kg IV) and maintained using isoflurane in 100% oxygen. A cranial
arthrotomy was performed on both femoropatellar joints.22 One femoropatellar
Acc ep ted Art icle
Microsoft Excel) to receive two critically-sized circular 15 mm diameter chondral defects on the medial trochlear ridge of the femur.18,22,23 A graphical depiction of study methodology is reported in Figure 1. One defect (of the two created) was randomly assigned to have calcified cartilage retained (Repair 1), while the second defect had calcified cartilage removed (Repair 2) to establish varying degrees of repair tissue in the defect joints. Subchondral bone microfracture was not
performed to avoid potential influence of the subchondral bone on assessments.24 The contralateral femoropatellar (control) joint underwent sham arthrotomy with no defects being created. After surgery, the incisions were closed, and horses were maintained on stall confinement until the incisions were healed 12 days after surgery. The horses were then released to a large paddock for free exercise.18 Four horses were sacrificed eight weeks and three horses sacrificed 16 weeks after defects were created using pentobarbital 86 mg/kg IV. After sacrifice, the stifles were frozen at -20°C.
Each stifle was thawed individually to accommodate the numerous in vitro assessments while preventing autolysis of tissues and multiple freeze-thaw cycles.
For each horse, the defect joint was processed first followed by the control joint to ensure biopsies were collected at matched sites between defect and control joints to address the inherent variability that exists across articular surfaces.25,26 After
thawing, osteochondral plug biopsies (7 mm diameter) were collected using a diamond tipped cylindrical coring drill bit (Starlite Industries, Bryn Mawr, PA)
Acc ep ted Art icle
attached to a drill press (Delta Power Equipment Company, Anderson, SC) under constant irrigation.19 In defect joints, 14 osteochondral cores were harvested; two within each defect (repair 1 and 2 sites), two immediately adjacent to each defect (adjacent sites), and two from the lateral trochlear ridge of the femur and four on the articular surface of the patella (remote sites) (Figure 1). These same 14 site- matched biopsies were collected from the control joints. In total, 196 osteochondral biopsy plugs were collected from all seven horses (14 joints). The articular cartilage from each biopsy was graded macroscopically on the ICRS scoring system.27 Each plug was allowed to equilibrate overnight with 0.9% saline in a preservative
solution containing non-specific protease inhibitors, antibiotics, and antimycotics (5 mM Benzamidine HCl, 5 mM EDTA both of Sigma-Aldrich, St. Louis, MO; 1x
Antibiotic-Antimycotic, Life Tech, Carlsbad, CA) at room temperature to regain isotonicity. Articular cartilage from each biopsy was analyzed for its mechanical, cationic CECT imaging, biochemical and histological characteristics.
Mechanical assessment
Prior to mechanical testing, articular cartilage thickness from each
osteochondral biopsy was determined with microCT (µCT40, Scanco Medical AG, Brüttisellen, Switzerland) using the following parameters 70 kVp, 113 µA, 300 ms integration time and 36-µm isotropic voxel resolution. The stress-relaxation compressive regimen was reported for 112 plugs (eight per joint). Plugs from
Acc ep ted Art icle
were analyzed. Repair 1 and 2 sites in the defect joint were initially tested though could not sustain the preload force (5 g, <0.13 MPa) and were excluded. The site- matched locations from repair 1 and 2 sites (healthy cartilage in control joints) were not mechanically evaluated to maintain sample group balance. Each analyzed plug was rigidly clamped in a mechanical testing apparatus (Enduratec3230, BOSE, Eden Prairie MN). A 5 N pre-load was applied to the articular cartilage surface in
unconfined compression using a nonporous ultrahigh molecular weight polyethylene platen and the sample was allowed to relax for 45 minutes while immersed in 400 mOsm/kg saline solution, replicating synovial fluid osmolality.28 The stress-relaxation regimen for each plug consisted of four incremental 5%
compressive strain steps (0.333% thickness/sec) with 45-minute stress relaxation between strain steps. Force and displacement data were collected at 10 Hz and a linear fit to stress versus strain at each equilibrium step was used to calculate compressive Young’s modulus (Eeq) using MATLAB (R2017a, Mathworks, Natick, MA). The dynamic response was calculated separately and reported at each displacement step. These data were used in calculating the initial equilibrium (E0) and strain-dependent (Eε) dynamic modulus.29 To estimate tissue permeability, a stretched exponential model (exp-(t/τ)^β) was fitted to all the steps from the experimental data, where τ is the stress-relaxation time constant and β the
stretching parameter.30 The τ and β values were averaged across steps and reported as a single value.
Acc ep ted Art icle
Cationic CECT assessment
All 196 osteochondral plugs underwent immersion in CA4+ (24 mg I/mL, 400 mOsmol/kg, pH=7.4) for 24 hours at 20°C to ensure equilibrium.10,31 Then, the plugs were imaged using microCT at the same above settings. The cationic CECT images were imported into Analyze® (v12.0, Biomedical Imaging Resource, Mayo Clinic, Rochester, MN). A semi-automatic threshold-based segmentation procedure was performed with manual correction to segment articular cartilage from the subchondral bone and air. Cationic CECT attenuation was converted to Hounsfield units (HUs) using a sample of deionized water and reported as a mean.
After imaging, the osteochondral plugs were equilibrated overnight in the preservative solution to remove residual CA4+.11,19 The articular cartilage of each biopsy was bisected and was removed from one side using a scalpel blade, while the remaining osteochondral portion was placed in 10% formalin for histologic analysis.
The portion of removed articular cartilage was minced and placed in a tube assigned for biochemical quantification. All samples were weighed to determine a hydrated (wet) weight. After lyophilization for 24 hours they were weighed again to
determine dry weight. Samples were stored at -80°C until further biochemical analysis.
Acc ep ted Art icle
Biochemical assessment
The biochemical content of articular cartilage from each osteochondral biopsy was determined using the 1,9-dimethylmethylene blue (DMMB) and hydroxyproline assays for GAG and total collagen content, respectively.2,32 The lyophilized cartilage was digested in papain (1 mg/mL, P4762, Sigma-Aldrich) at 65°C. The DMMB assay was performed using a standard curve generated with known concentrations of chondroitin C sulfate (C6737, Sigma-Aldrich, St. Louis, MO). Samples were prepared in triplicate and read on a microplate reader
(SpectraMax M3, Molecular Devices, Sunnyvale, CA), set at a wavelength of 530 nm.
Samples were repeated if the standard curve R2 was < 0.95 or if the coefficient of variation between replicates was > 0.1. Mean GAG concentration of each sample was standardized on a % wet weight basis. For determination of total collagen content, the sample was hydrolyzed with an equal volume of 12.1 N HCl for 16 hours at 110°C and evaporated on a heating block set at 60°C overnight. Samples along with known hydroxyproline standard concentrations (H5534, Sigma-Aldrich) were plated in duplicate. Chloramine T reagent (50 mM, 857319, Sigma-Aldrich) was added to the samples and hydroxyproline standards, which were incubated for 20 minutes at 25°C. The 4-dimethyl-aminobenzaldehyde reagent (1M, 156477, Sigma- Aldrich) was added and incubated at 60°C for 15 minutes followed by 5 minutes at 25°C. Then, the plate was read at 550 nm. Hydroxyproline concentrations for each sample were derived from the standard curve. Total collagen was determined using
Acc ep ted Art icle
an established conversion (7.6 mg hydroxyproline/mg collagen),33 reported as a mean and standardized to dry weight.
Histological assessment
After formalin fixation, plugs were decalcified (Formical-2000, Statlab, McKinney, TX), embedded in paraffin and five μm microscope slides were prepared.
Osteochondral samples were stained with hematoxylin and eosin and safranin-O fast green (SOFG). Control tissues (bovine trachea and equine osteochondral
samples) were stained concurrently to ensure consistency across batches. Articular cartilage was scored using the ICRS II system to characterize reparative articular cartilage.34 Three investigators scored a random subset (30 samples) of the slides.
Once grading was consistent between the three scorers, all slides were anonymized and randomized. Then, a single grader performed all ICRS II scoring assessments.
Briefly, each component was graded on a continuous scale 0 – 100 (0 = poor, 100 = excellent).34 An ICRS II composite score was created by averaging the individual scoring components. Regional SOFG matrix staining (superficial, middle and deep territorial and interterritorial zones) was also scored separately.
Acc ep ted Art icle
Data and statistical analysis
Descriptive statistics of continuous data were reported as mean ± s.d..
Categorical data were reported as median and range. Correlations between variables were calculated accounting for repeated measures on subjects and correlation strength characterized (slight: 0.0 to 0.20, fair: 0.21 to 0.40, moderate:
0.41 to 0.60, strong: 0.61 to 0.80, very strong: 0.81 to 1.0).35,36 Osteochondral plug data were compared using a mixed model analysis. Fixed effects included joint (defect or control), site (repair 1, repair2, adjacent, remote) and joint*site
interaction. Random effects were horse and horse*joint interaction to account for repeated measures. Tukey Kramer adjustments were used to reduce type I error for multiple comparisons. Assumptions of each statistical assessment were verified based on visual inspection of residual diagnostic plots. Mechanical, microCT, biochemical and histologic outcomes were not different between the two sacrifice time points. Therefore, sacrifice time point was not included as a covariate.
Statistical analyses were performed using SAS (SAS University Edition, v9.2, SAS Institute Inc., Cary, NC) and significance was defined at P<0.05.
Results
Repair 1 defects had minimal and irregular amounts of tissue filling (ICRS score median: 3, range: 3 – 4), while repair 2 defects had complete filling of repair tissue (ICRS score median: 2, range: 2 – 3) (Figure 2). The perimeter surrounding both circular defects was irregular and visually different from the sharp edges
Acc ep ted Art icle
observed at the end surgery; indicative of peri-defect deterioration. Adjacent (to defect) sites had near normal macroscopic appearance (ICRS score median: 0, range:
0 – 1). The remote sites in the defect joints appeared healthy (ICRS score median: 0, range: 0 – 0). All sites within the control joints had a confluent, normal appearing articular cartilage surface (ICRS score median: 0, range: 0 – 0).
Cationic CECT assessment
The distribution of CA4+ throughout articular cartilage varied based on disease state. Mean cationic CECT attenuation was significantly different between joint sites (P<0.0001). Cationic CECT attenuation was significantly lower in repair 1 (1478 ± 333 HUs), repair 2 (1229 ± 191 HUs) and adjacent sites (2139 ± 336 HUs) in defect joints when compared to their matched sites in control joints (2587 ± 298, 2505 ± 184 and 2563 ± 538 HUs, respectively; all P<0.0001) (Figure 2). Within defect joints, the mean cationic CECT attenuation was significantly higher at the remote site (2928 ± 420 HUs) when compared to repair 1, repair 2 and adjacent sites (all P<0.0001). A significant difference of cationic CECT attenuation between repair 1 and repair 2 sites in defect joints was not detected.
Acc ep ted Art icle
Mechanical assessment
Due to the preload overwhelming repair 1 and repair 2 tissue, only
mechanical data from adjacent and remote sites in defect and control joints were reported. Young’s equilibrium modulus (Eeq)was significantly different between joint sites (P=0.01). In the defect joints, mean Eeq at adjacent sites (0.33 ± 0.16 MPa) was significantly lower than at remote sites (0.58 ± 0.20 MPa, P<0.0001) (Figure 3).
The mean Eeq at adjacent sites (in defect joints) was significantly lower than matched sites in control joints (0.50 ± 0.14, P=0.01) (Figure 3). Mean τ was
significantly different between joint sites (P=0.004). In the defect joints, mean τ at adjacent sites (70 ± 18 s) was significantly lower than remote sites (132 ± 33 s, P<0.0001) (Figure 3). The mean τ at adjacent sites (in defect joints) was
significantly lower than matched sites in control joints (143 ± 29 s, P=0.003). Mean β was significantly different between joint sites (P<0.0001). In the defect joints, mean β at adjacent sites (0.44 ± 0.02) was significantly lower than remote sites (0.49 ± 0.03, P<0.0001) (Figure 3). The mean β at adjacent sites (in defect joints) was significantly lower than matched sites in control joints (0.50 ± 0.02, P=0.0008).
Cationic CECT strongly correlated with Eeq (r = 0.71, P<0.0001), τ (r = 0.68,
P<0.0001) and β (r = 0.72, P<0.0001), and moderately correlated with Eε (r = 0.52, P<0.0001) and E0 (r = 0.50, P<0.0001) (Figure 3 and 4, Table 1).
Acc ep ted Art icle
Biochemical assessment
Mean GAG content varied by joint site (P<0.0001). GAG was significantly lower in repair 1 (1.7 ± 2.0%) and repair 2 (0.5 ± 0.2%) sites in defect joints when compared to their matched sites in control joints (3.9 ± 1.5%, P=0.0004; and 3.8 ± 0.9%, P<0.0001; respectively) (Figure 5). Within defect joints, GAG content was significantly higher at the remote site (4.1 ± 1.2%) when compared to repair 1, repair 2 and adjacent sites (all P<0.0001). A significant difference in GAG between repair 1 and repair 2 sites in defect joints was not detected. However, GAG was significantly higher at adjacent sites (3.2% ± 1.1%) than repair 2 sites in defect joints (P<0.0001). Cationic CECT strongly correlated with GAG (r = 0.755, P<0.0001). Mean collagen content varied by joint sites (P=0.028). Within defect joints, collagen content was significantly higher in remote sites (56 ± 22%) than in repair 1 (38 ± 22%, P=0.018) and repair 2 (45 ± 30%, P=0.05) sites. A difference in collagen between adjacent and remote sites was not detected. Cationic CECT slightly correlated with collagen content (P=0.05).
Histological assessment
Mean ICRS II composite score varied by joint site (P<0.0001). Mean ICRS II composite score was significantly lower in repair 1 (48.6 ± 6.9), repair 2 (40.0 ± 5.4) and adjacent sites (80.3 ± 7.5) in defect joints when compared to matched sites in control joints (88.5 ± 4.4, 90.7 ± 3.2 and 89.3 ± 4.2, respectively; all P<0.0001). In
Acc ep ted Art icle
sites (all P<0.0001), except no difference was detected between repair 1 and repair 2 sites (Figure 6). Cationic CECT was strongly correlated with ICRS II composite score (r = 0.794, P<0.0001) and with most ICRS II individual scoring components (Table 1). Comparisons of cationic CECT with ICRS II scoring components between defect and control joints are depicted in Supplementary Item 1.
Discussion
Evaluation of early articular cartilage injury is a diagnostic challenge and innovative imaging strategies to characterize disease states are needed. Cationic CECT, through its ability to proportionally diffuse into articular cartilage matrix based on the GAG fixed negative charge density, enables assessment of cartilage and potentially the detection and monitoring of articular cartilage injury and repair.
Using an equine translational model, cationic CECT discriminates different states of articular cartilage health in a continuum of articular cartilage injury and repair states. In support of our hypotheses, cationic CECT successfully differentiates between reparative, degenerative and healthy articular cartilage. Additionally, cationic CECT attenuation highly correlates with biochemical, mechanical and histologic parameters.
While studies have examined cationic CECT in degenerative cartilage,11,19,31 this study shows that reparative tissue can be segregated from early degenerative
Acc ep ted Art icle
(adjacent sites) and healthy (remote sites) articular cartilage using in vivo generated tissue. Notable differences exist between the repair groups. Repair 1 tissue retains calcified cartilage in the defect. This technique is not used clinically because of the minimal tissue ingrowth with inferior tissue quality that forms,24,37 as was
corroborated by this study. Because we performed gentle debridement to retain calcified cartilage in repair 1 defects, some noncalcified deep articular cartilage may have remained,38 though it was not detected on histologic review. The repair 2 group imitates calcified cartilage removal as would be performed prior to
microfracture and marrow stimulation procedures used clinically.24,37 The resultant tissue is fibrous to fibrocartilaginous on histological assessment. Repair 2 tissue more completely fills the defect than repair 1 tissue but is still soft and of inferior quality when compared to adjacent and remote sites. In both repair groups, cationic CECT attenuation is lower than remote and adjacent sites because proteoglycan concentrations (and negative fixed charge density) are not reinstituted during the healing process. Thus, CA4+ diffusion and retention are minimal. Interestingly, the adjacent site samples showed a gradient of degradation histologically consistent with the distribution of cationic CECT attenuation (Figure 2). This radial
degeneration is known to occur with this equine model,20 and recapitulates the need for cartilage repair techniques to reinstitute the mechanical strength and
congruency of the entire articular surface.
Acc ep ted Art icle
Cationic CECT attenuation values highly correlate with biochemical, mechanical and histological properties; establishing a method that distinguishes articular cartilage quality based on internal composition as well as structural and mechanical attributes. Several studies report significant correlations between cationic CECT with GAG11-15 and mechanical properties13-15in vitro, and this study shows that these findings also apply with in vivo generated repair tissue.
Notwithstanding the discrimination of reparative from degenerative and normal tissue, cationic CECT is unable to distinguish between the repair 1 and 2 tissue groups. This lack of differentiation likely manifests as a consequence of the study design as no composite or individual ICRS II histologic scoring parameters were capable of distinguishing between these two repair groups either. The lack of
biochemical and histological differences between repair 1 and 2 tissue are likely due to the short-term nature of the study but also reflects the lack of microfracture or other marrow-stimulating procedures to improve repair tissue quality. Cationic CECT attenuation represents total tissue attenuation irrespective of tissue volume.
Whereas tissue volume can affect contrast media diffusion (due to a varying distribution of constituents in healthy articular cartilage), the homogenous
fibrous/fibrocartilaginous tissue in these defects did not change in composition with increased tissue volume. However, cationic CECT does capture repair tissue volume at high resolution. Thus, future studies should consider volume quantification of repair tissue in addition to reporting cationic CECT attenuation to provide a comprehensive tissue assessment and a potential way to differentiate between repair 1 and repair 2 tissue.
Acc ep ted Art icle
Cationic CECT attenuation mirrors most of the ICRS II histological scoring components. Regardless, articular cartilage is a heterogeneous tissue and
classification of its health from a conglomerate score could impart inaccuracy. While regional properties are indirectly appreciated with cationic CECT, a segmental tissue assessment using cationic CECT could be pursued to obtain higher accuracy depth-dependent characterizations. Similar to a histologic scoring system for use in posttraumatic osteoarthritic cartilage,19,39 the amount of SOFG staining in the ICRS II score in this study highly relates to cationic CECT attenuation owing to the similar mechanism of action of cationic molecules on GAG charge.19,40 While the reparative groups do not take in safranin-O (a cationic dye) stain, CA4+ does diffuse into these same tissues. This disparity is attributed to differences between microCT imaging and histologic methods. MicroCT provides a volumetric assessment, whereas histology is based on planar assessment. Because of the tissue volume in the intact plug, hydrated solutions (and some CA4+) are passively drawn in irrespective of low charge.9,41,42 In contrast, the safranin-O stain is only attracted to negative charges or is washed away during processing. Intuitively, cationic CECT attenuation reflects GAG content due to electrostatic interactions; however, additional correlations with other parameters outside of the extracellular matrix (e.g. chondrocytes) are less clear. While chondrocytes synthesize GAGs, chondrocyte loss in disease typically echoes the depletion of matrix constituents.43 Considering the overlap that exists across individual histologic scoring parameters and their interaction with one
Acc ep ted Art icle
another (co-linearity) throughout the disease process, the correlation of cationic CECT attenuation with constituents outside the extracellular matrix are likely because of those indirect influences.
Typically, this translational model is used for preclinical investigation of cartilage healing techniques in humans.18,22,23,44 While the same surgical technique was replicated here, the short duration of study is not considered appropriate for the investigation of long-term healing.18 This study was designed to examine the capability of cationic CECT to detect articular cartilage injury and characterize tissue quality across a variety of health states. Further work will be required to examine the utility of this method long-term and to serially monitor progressive healing in vivo. Despite the categorization as reparative tissue, the methods used for healing in this study are not the optimal techniques to maximize cartilage repair in clinical practice today. Subchondral bone microfracture and osteochondral allograft transplantation used in horses and humans would certainly lead to better quality repair tissue.24,45 Repair tissue after subchondral bone microfracture is
fibrocartilaginous and osteochondral transplantation re-institutes native hyaline cartilage with clear mechanical and biochemical differences between these tissue types.46-49 While cationic CECT attenuation is expected to distinguish between fibrocartilage and reparative hyaline cartilage because of its mechanism of action, further experiments are required to verify this hypothesis. Image resolution is higher on microCT than scanners used in a clinical setting. In this current study,
Acc ep ted Art icle
voxel dimensions of the microCT was 36-µm, while clinical scanners have approximately 10-fold lower in-plane resolution. However, this difference in resolution does not preclude estimations of GAG content from clinical scanners.19,50 Studies investigating the capacity for cationic CECT to monitor healing
progression,18,23 and how those outcomes compare to quantitative MRI assessments in MRI and CT scanners in clinical settings,6,8,9 are necessary.
In conclusion, cationic CECT provides imaging biomarkers that reflect tissue quality through non-destructive assessment and successfully distinguishes between reparative, degenerative and healthy articular cartilage. Further investigation into the ability for cationic CECT to be used for in vivo assessmentin clinical scanners warrants exploration. These data from equine translational models would establish the feasibility of investigating these methods in human tissues. This quantitative imaging technique is a promising investigative tool for longitudinal assessment of articular cartilage and as a method to characterize articular cartilage injury across a variety of disease states.
Acknowledgments
Funding for this project was obtained from the Grayson Jockey Club Research Foundation, College of Veterinary Medicine and Biomedical Sciences Cooperative
Acc ep ted Art icle
Foundation, Päivikki and Sakari Sohlberg Foundation, and Orion Research
Foundation (Mäkelä). The authors acknowledge Nikki Phillips and Dr. John Kisiday for assistance with biochemical analyses, and Jen Daniels, Dr. Kelly Zersen, Dr. Katie Seabaugh and the animal care staff at the Orthopaedic Research Center for their care of the horses.
References
1. McIlwraith CW, Frisbie DD, Kawcak CE. 2012. The horse as a model of naturally occurring osteoarthritis. Bone Joint Res 1: 297-309.
2. Frisbie DD, Al-Sobayil F, Billinghurst RC, et al. 2008. Changes in synovial fluid and serum biomarkers with exercise and early osteoarthritis in horses.
Osteoarthritis Cartilage 16: 1196-1204.
3. Bekkers JE, Creemers LB, Dhert WJ, et al. 2010. Diagnostic modalities for diseased articular cartilage-from defect to degeneration: a review. Cartilage 1: 157- 164.
4. Roemer FW, Eckstein F, Hayashi D, et al. 2014. The role of imaging in osteoarthritis. Best Pract Res Clin Rheumatol 28: 31-60.
5. Bansal PN, Joshi NS, Entezari V, et al. 2010. Contrast enhanced computed
tomography can predict the glycosaminoglycan content and biomechanical properties of articular cartilage. Osteoarthritis Cartilage 18: 184-191.
Acc ep ted Art icle
6. Taylor C, J Carballido-Gamio, Majumdar S, et al. 2009. Comparison of quantitative imaging of cartilage for osteoarthritis: T2, T1ρ, dGEMRIC and contrast- enhanced computed tomography. Mag Res Imag 27: 779-784.
7. Potter HG, Black BR, Chong le R. 2009. New techniques in articular cartilage imaging. Clin Sports Med 28: 77-94.
8. Hontoir F, Clegg P, Nisolle JF, et al. 2015. Magnetic resonance compositional imaging of articular cartilage: What can we expect in veterinary medicine?
Vet J 205: 11-20.
9. Nelson BB, Kawcak CE, Barrett MF, et al. 2018. Recent advances in articular cartilage evaluation using computed tomography and magnetic resonance imaging. Equine Vet J 50: 564-579.
10. Joshi NS, Bansal PN, Stewart RC, et al. 2009. Effect of contrast agent charge on visualization of articular cartilage using computed tomography: exploiting electrostatic interactions for improved sensitivity. J Am Chem Soc 131:
13234-13235.
11. Bansal PN, Joshi NS, Entezari V, et al. 2011. Cationic contrast agents improve quantification of glycosaminoglycan (GAG) content by contrast enhanced CT imaging of cartilage. J Orthop Res 29: 704-709.
12. Bansal PN, Stewart RC, Entezari V, et al. 2011. Contrast agent electrostatic
Acc ep ted Art icle
contrast uptake ratio and improved quantitative CT imaging in cartilage.
Osteoarthritis Cartilage 19: 970-976.
13. Lakin BA, Ellis DJ, Shelofsky JS, et al. 2015. Contrast-enhanced CT facilitates rapid, non-destructive assessment of cartilage and bone properties of the human metacarpal. Osteoarthritis Cartilage 23: 2158-2166.
14. Lakin BA, Grasso DJ, Shah SS, et al. 2013. Cationic agent contrast-enhanced computed tomography imaging of cartilage correlates with the compressive modulus and coefficient of friction. Osteoarthritis Cartilage 21: 60-68.
15. Lakin BA, Patel H, Holland C, et al. 2016. Contrast-enhanced CT using a cationic contrast agent enables non-destructive assessment of the biochemical and biomechanical properties of mouse tibial plateau cartilage. J Orthop Res 34:
1130-1138.
16. Kallioniemi AS, Jurvelin JS, Nieminen MT, et al. 2007. Contrast agent enhanced pQCT of articular cartilage. Phys Med Biol 52: 1209-1219.
17. Kulmala KAM, Karjalainen HM, Kokkonen HT, et al. 2013. Diffusion of ionic and non-ionic contrast agents in articular cartilage with increased cross-linking-- contribution of steric and electrostatic effects. Med Eng Phys 35: 1415-1420.
18. McIlwraith CW, Fortier LA, Frisbie DD, et al. 2011. Equine models of articular cartilage repair. Cartilage 2: 317-326.
Acc ep ted Art icle
19. Nelson BB, Mäkelä JTA, Lawson TB, et al. 2019. Evaluation of equine articular cartilage degeneration after mechanical impact injury using cationic
contrast-enhanced computed tomography. Osteoarthritis Cartilage 27: 1219- 1228.
20. Strauss EJ, Goodrich LR, Chen CT, et al. 2005. Biochemical and biomechanical properties of lesion and adjacent articular cartilage after chondral defect repair in an equine model. Am J Sports Med 33: 1647-1653.
21. American Association of Equine Practitioners. Guide for veterinary service and judging of equestrian events. Lexington, KY: American Association of Equine Practitioners, 1991.
22. Frisbie DD, Bowman SM, Colhoun HA, et al. 2008. Evaluation of autologous chondrocyte transplantation via a collagen membrane in equine articular defects – results at 12 and 18 months. Osteoarthritis Cartilage 16: 667-679.
23. Hurtig MB, Fretz PB, Doige CE, et al. 1988. Effects of lesion size and location on equine articular cartilage repair. Can J Vet Res 52: 137-146.
24. Frisbie DD, Trotter GW, Powers BE, Rodkey WG, Steadman JR, Howard RD, Park RD, and McIlwraitth CW. 1999. Arthroscopic subchondral bone plate
microfracture technique augments healing of large chondral defects in the radial carpal bone and medial femoral condyle of horses. Vet Surg 28: 242- 255.
Acc ep ted Art icle
25. Frisbie DD, Cross MW, McIlwraith CW. 2006. A comparative study of articular cartilage thickness in the stifle of animal species used in human pre-clinical studies compared to articular cartilage thickness in the human knee. Vet Comp Orthop Traumatol 19: 142-146.
26. Changoor A, Hurtig MB, Runciman RJ, et al. 2006. Mapping of donor and recipient site properties for osteochondral graft reconstruction of
subchondral cystic lesions in the equine stifle joint. Equine Vet J 38: 330-336.
27. Brittberg M, Winalski CS. 2003. Evaluation of cartilage injuries and repair. J Bone Joint Surg Am 85-A Suppl 2: 58-69.
28. Urban JP, Hall AC, Gehl KA. 1993. Regulation of matrix synthesis rates by the ionic and osmotic environment of articular chondrocytes. J Cell Physiol 154:
262-270.
29. Huttu MR, Puhakka J, Makela JT, et al. 2014. Cell-tissue interactions in
osteoarthritic human hip joint articular cartilage. Connect Tissue Res 55:
282-291.
30. June RK, Mejia KL, Barone JR, et al. 2009. Cartilage stress-relaxation is affected by both the charge concentration and valence of solution cations.
Osteoarthritis Cartilage 17: 669-676.
31. Nelson BB, Stewart RC, Kawcak CE, et al. 2018. Quantitative evaluation of equine articular cartilage using cationic contrast-enhanced computed tomography.
Acc ep ted Art icle
Cartilage [Epub ahead of print]
https://doi.org/10.1177/1947603518812562.
32. Lippiello L, Hall D, Mankin HJ. 1977. Collagen synthesis in normal and osteoarthritic human cartilage. J Clin Invest 59: 593-600.
33. Venn M, Maroudas A. 1977. Chemical composition and swelling of normal and osteoarthrotic femoral head cartilage. I. Chemical composition. Ann Rheum Dis 36: 121-129.
34. Mainil-Varlet P, Van Damme B, Nesic D, et al. 2010. A new histology scoring system for the assessment of the quality of human cartilage repair: ICRS II.
Am J Sports Med 38: 880-890.
35. Bland JM, Altman DG. 1995. Calculating correlation coefficients with repeated observations: Part 1--Correlation within subjects. BMJ (Clinical research ed) 310: 446-446.
36. Landis JR, Koch GG. 1977. The measurement of observer agreement for categorical data. Biometrics 33: 159-174.
37. Gelse K, Riedel D, Pachowsky M, et al. 2015. Limited integrative repair capacity of native cartilage autografts within cartilage defects in a sheep model. J Orthop Res 33: 390-397.
38. Hoemann CD, Lafantaisie-Favreau CH, Lascau-Coman V, et al. 2012. The
Acc ep ted Art icle
39. McIlwraith CW, Frisbie DD, Kawcak CE, et al. 2010. The OARSI histopathology initiative - recommendations for histological assessments of osteoarthritis in the horse. Osteoarthritis Cartilage 18: S93-S105.
40. Schmitz N, Laverty S, Kraus VB, et al. 2010. Basic methods in histopathology of joint tissues. Osteoarthritis Cartilage 18: S113-S116.
41. Camplejohn KL, Allard SA. 1988. Limitations of safranin 'O' staining in
proteoglycan-depleted cartilage demonstrated with monoclonal antibodies.
Histochemistry 89: 185-188.
42. Maroudas A. 1976. Balance between swelling pressure and collagen tension in normal and degenerate cartilage. Nature 260: 808-809.
43. van Weeren PR. 2016. General anatomy and physiology of joints. In: McIlwraith CW, Frisbie DD, Kawcak CE, et al., editors. Joint disease in the horse, 2nd ed.
St. Louis, MO: Elsevier; p 1-24.
44. Kol A, Arzi B, Athanasiou KA, et al. 2015. Companion animals: Translational scientist's new best friends. Sci Transl Med 7: 1-8.
45. Torrie AM, Kesler WW, Elkin J, et al. 2015. Osteochondral allograft. Curr Rev Musculoskelet Med 8: 413-422.
46. Valiyaveettil M, Mort JS, McDevitt CA. 2005. The concentration, gene expression, and spatial distribution of aggrecan in canine articular cartilage, meniscus, and anterior and posterior cruciate ligaments: a new molecular distinction
Acc ep ted Art icle
between hyaline cartilage and fibrocartilage in the knee joint. Connect Tissue Res 46: 83-91.
47. Nguyen AM, Levenston ME. 2012. Comparison of osmotic swelling influences on meniscal fibrocartilage and articular cartilage tissue mechanics in
compression and shear. J Orthop Res 30: 95-102.
48. LeRoux MA, Arokoski J, Vail TP, et al. 2000. Simultaneous changes in the
mechanical properties, quantitative collagen organization, and proteoglycan concentration of articular cartilage following canine meniscectomy. J Orthop Res 18: 383-392.
49. Silver FH, Bradica G, Tria A. 2001. Relationship among biomechanical, biochemical, and cellular changes associated with osteoarthritis. Crit Rev Biomed Eng 29: 373-391.
50. Honkanen MKM, Matikka H, Honkanen JTJ, et al. 2019. Imaging of proteoglycan and water contents in human articular cartilage with full-body CT using dual contrast technique. J Orthop Res 37: 1059-1070.
Tables
Table 1 – Correlations between cationic contrast-enhanced computed tomography (CECT) attenuation with biochemical, mechanical and histological parameters in equine articular cartilage across disease states. Correlation coefficients (r) and P-
Acc ep ted Art icle
correction controlling for repeated measures on subjects. GAG, glycosaminoglycan;
Eeq, equilibrium compressive Young’s modulus; Eε, strain-dependent dynamic
modulus; E0, initial dynamic modulus; Edyn, dynamic Young’s modulus at each testing step; τ, stress-relaxation time constant; β, stretching parameter constant; ICRS, International Cartilage Repair Society; SOFG, safranin-O fast green. The ICRS II composite score is the average of all components within the scoring system.
Cationic CECT Attenuation
Parameter r P-value
GAG concentration 0.755 <0.0001
Eeq 0.710 <0.0001
Eε 0.523 <0.0001
E0
0.497 <0.0001
5% Edyn 0.603 <0.0001
10% Edyn 0.578 <0.0001
Acc ep ted Art icle
15% Edyn 0.578 <0.0001
20% Edyn 0.586 <0.0001
τ 0.681 <0.0001
β 0.724 <0.0001
Total collagen 0.171 0.05
ICRS II composite score 0.794 <0.0001
Tissue morphology score 0.701 <0.0001
Matrix staining score 0.697 <0.0001
Cell morphology score 0.722 <0.0001
Chondrocyte clustering score 0.424 <0.0001
Surface architecture score 0.717 <0.0001
Acc ep ted Art icle
Basal integration score 0.654 <0.0001
Tidemark formation score 0.726 <0.0001
Subchondral bone abnormality score 0.729 <0.0001
Inflammation score -0.071 0.33
Abnormal calcification ossification
score 0.350
<0.0001
Vascularization in repair score 0.488 <0.0001
Surface superficial assessment score 0.740 <0.0001
Mid deep assessment score 0.756 <0.0001
Overall assessment 0.765 <0.0001
SOFG superficial score 0.357 <0.0001
SOFG middle score 0.708 <0.0001
Acc ep ted Art icle
SOFG deep territorial score 0.463 <0.0001
SOFG deep interterritorial score 0.554 <0.0001
ICRS score -0.742 <0.0001
Acc ep ted Art icle
Figure Legends
Figure 1: Summary of study methodology. Two critically-sized chondral defects were created on the medial trochlear ridge of the femur and randomized to have calcified cartilage retained (repair 1) or removed (repair 2). At sacrifice
osteochondral plugs were collected across the articular surface of the femoral trochlea and patella at 4 sites (repair 1, repair 2, adjacent to defects or at remote sites). Plugs were scored macroscopically with the International Cartilage Repair Society (ICRS) scale and subsequently underwent mechanical testing, cationic micro CT imaging, biochemical quantification and histological assessment using the ICRS II score.
Acc ep ted Art icle
Figure 2: Comparison of cationic contrast-enhanced computed tomography (CECT) attenuation between reparative (repair 1 and 2), degenerative (adjacent) and healthy (remote) articular cartilage within defect and control joints (top image).
Data are mean ± standard deviation. Connecting bars indicate a significant
difference between groups. Images below groups represent cationic CECT, safranin- O histology and arthroscopic images of tissue samples within a defect joint.
Acc ep ted Art icle
Figure 3: Correlation of cationic CECT attenuation with equilibrium compressive modulus (Eeq) (left image). Data points represent each plug analyzed and are coded by macroscopic ICRS score. Correlations were made using Bland’s correction to account for repeated measures on subjects. Repair tissue from repair 1 and 2 groups was not strong enough to sustain compressive preloads and were excluded from mechanical assessments. Comparison of Eeq at sites adjacent and remote to defects in defect and control joints (right image). Data are represented as mean ± standard deviation. Connecting bars indicate a significant difference between groups.
Acc ep ted Art icle
Figure 4: Correlation of cationic CECT attenuation with tissue permeability stress- relaxation time (τ, left image) and stretching parameter (β, middle image) constants, and strain-dependent dynamic modulus (Eε, right image). Data points represent each plug analyzed and are coded by macroscopic ICRS score. Correlations were made using Bland’s correction to account for repeated measures on subjects. Repair tissue from repair 1 and 2 groups was not strong enough to sustain compressive preloads and were excluded from mechanical assessments.
Acc ep ted Art icle
Figure 5: Correlation of cationic CECT attenuation with glycosaminoglycan (GAG) content (left image). Data points represent each plug analyzed and are coded by macroscopic ICRS score. Correlations were made using Bland’s correction to
account for repeated measures on subjects. Comparison of GAG at sites of repair and sites adjacent and remote to defects in defect and control joints (right image). Data are represented as mean ± standard deviation. Connecting bars indicate a significant difference between groups.
Acc ep ted Art icle
Figure 6: Correlation of cationic CECT attenuation with International Cartilage Repair Society (ICRS) II histologic composite score (left image). Data points represent each plug analyzed and are coded by macroscopic ICRS score.
Correlations were made using Bland’s correction to account for repeated measures on subjects. Comparison of GAG at repair sites and sites adjacent and remote to defects in defect and control joints (right image). Data are represented as mean ± standard deviation. Connecting bars indicate a significant difference between groups.
