Effects of human articular cartilage constituents on simultaneous diffusion of cationic and nonionic contrast agents

Abstract Contrast‐enhanced computed tomography is an emerging diagnostic technique for osteoarthritis. However, the effects of increased water content, as well as decreased collagen and proteoglycan concentrations due to cartilage degeneration, on the diffusion of cationic and nonionic agents, are not fully understood. We hypothesize that for a cationic agent, these variations increase the diffusion rate while decreasing partition, whereas, for a nonionic agent, these changes increase both the rate of diffusion and partition. Thus, we examine the diffusion of cationic and nonionic contrast agents within degraded tissue in time‐ and depth‐dependent manners. Osteochondral plugs (N = 15, d = 8 mm) were extracted from human cadaver knee joints, immersed in a mixture of cationic CA4+ and nonionic gadoteridol contrast agents, and imaged at multiple time‐points, using the dual‐contrast method. Water content, and collagen and proteoglycan concentrations were determined using lyophilization, infrared spectroscopy, and digital densitometry, respectively. Superficial to mid (0%‐60% depth) cartilage CA4+ partitions correlated with water content (R < −0.521, P < .05), whereas in deeper (40%‐100%) cartilage, CA4+ correlated only with proteoglycans (R > 0.671, P < .01). Gadoteridol partition correlated inversely with collagen concentration (0%‐100%, R < −0.514, P < .05). Cartilage degeneration substantially increased the time for CA4+ compared with healthy tissue (248 ± 171 vs 175 ± 95 minute) to reach the bone‐cartilage interface, whereas for gadoteridol the time (111 ± 63 vs 179 ± 163 minute) decreased. The work clarifies the diffusion mechanisms of two different contrast agents and presents depth and time‐dependent effects resulting from articular cartilage constituents. The results will inform the development of new contrast agents and optimal timing between agent administration and joint imaging.


| INTRODUCTION
Articular cartilage is avascular, and its metabolic function is regulated via diffusion and convection of charged and uncharged solutes between the synovial fluid and the constituents of the cartilage extracellular matrix (ECM). 1 Cartilage ECM is a heterogeneous structure, mainly consisting of interstitial water (60%-85%), collagen fibrils (50%-80% of dry content), and negatively charged proteoglycans (PGs; 20%-30% of dry content). 2,3 Changes in the tissue composition alter the interstitial fluid flow 2,4 and mechanical properties. [5][6][7] The diffusion of a contrast agent inside the tissue, followed by subsequent contrast-enhanced imaging, provides information on the health status of the cartilage tissue. [8][9][10][11][12] For example, contrastenhanced computed tomography (CECT) is used to evaluate osteoarthritis (OA)-related degeneration of cartilage and the associated alterations in the composition and morphology. [13][14][15][16] CECT diffusion studies of articular cartilage typically employ a single contrast agent. 11,13,14,17 In diffusion equilibrium, the partition of a nonionic agent follows the depth-wise profile of the interstitial water content. 3,15 However, since OA-related degeneration of cartilage affects all cartilage constituents, as well as the structure, sensitive quantification of cartilage health based on the partition of only a nonionic agent, is challenging. Anionic agents similarly suffer from low sensitivity, as they diffuse against the fixed negative charge that prevails inside healthy articular cartilage. In contrast, cationic contrast agents molecules are attracted into the tissue through electrostatic attraction and are used to directly quantify cartilage PG concentration. 12,14,18 Unhealthy articular cartilage possesses a disorganized collagen fibril network and increased permeability; thus facilitating agent diffusion. 19,20 However, the fixed charge density is concurrently reduced, because of the decrease in PG concentration, which slows down the diffusion of cationic agents. The combination of the two simultaneous and opposite effects complicates the interpretation of the acquired results, which in turn reflects the overall tissue health.
To address this challenge, we recently introduced a dual-contrast agent technique. In this technique, two CT-based contrast agents (iodine I-based cationic, CA4+) 21 and gadolinium Gd-based nonionic agent [gadoteridol]) are employed simultaneously ( Figure 1) and the molar concentrations of the agents are quantified using a dualenergy CT scan. 12,22,23 The premise is that normalization of the cationic contrast agent partition with that of the nonionic contrast agent allows early diagnostics, as the changes in the tissue's steric hindrance are accounted for. The dual-contrast method shows improved sensitivity and assessment of cartilage properties. 12,22,23 However, questions still remain regarding the effects of the cartilage constituents and its hierarchical structure on the diffusion, for example, how the contrast agent flux in the superficial zone of cartilage differs from that in the deep cartilage, and how agent diffusion relates to the variation in the depth-wise organization of the cartilage constituents?
In this study, we characterized the effects of the main cartilage constituent content, that is, PGs, collagen, and water, and their changes during OA-related cartilage degradation, on the simultaneous diffusion of cationic and nonionic contrast agents. We evaluated the composition of the human articular cartilage samples via microscopy and spectroscopy and measured the diffusion of the contrast agents by dual-contrast CECT. The molecular size of gadoteridol was measured with a freely available open-source web-application to be~11 Å long and~6 Å wide (MolView, 2015). 25 The osmolality of the contrast agent bath was selected to be similar to physiological saline, which is safe for clinical application. 26 The bath was supplemented with following proteolytic During the immersion in the contrast agent, the baths were constantly stirred and kept at a temperature of 4°C. contrast agent diffusion curves were determined for 20% thick sections (0%-20%, 20%-40%, 40%-60%, 60%-80%, and 80%-100% of cartilage depth) by fitting the following equation to the diffusion

| Image analysis
where C max is the contrast agent concentration maximum, t is the diffusion time, and T is the time required for the contrast agent to reach 63.2% of the maximum concentration. 18 The diffusion of the contrast agents was examined separately for five 20% thick cartilage sections with a partition threshold of 20%. This threshold was chosen to ensure sufficient temporal and spatial resolutions for determination of the contrast agent diffusion times.

| Reference methods
Water content measurements were carried out on the osteochon- light source (λ = 420 ± 5 nm) and a 12-bit CCD (ORCA-ER, Hamamatsu Photonics K.K., Japan). 29 Before the DD measurements, the system was calibrated using neutral density filters (Schott, Germany) with an OD range between 0 and 3. From the DD measurements, depth-wise OD profiles from the cartilage surface to cartilage-bone interface were calculated ( Figure 2K).
Collagen concentration distribution was determined using

| Statistical analysis
To evaluate the effect of cartilage degeneration on contrast agent diffusion the samples were grouped based on the Mankin score ("more degenerated": Mankin score > 5, n = 8, average score = 6.9 Histological analyses showed that PG and amide I concentrations predominantly increased while water content decreased as a function of cartilage depth ( Figure 3). 30,[33][34][35] The differences between the distributions of collagen (amide I) and PG concentration between more (Mankin score > 5) and less degenerated samples (Mankin score ≤ 5) were not statistically significant.

| Diffusion as a function of cartilage depth
The rate of diffusion was similar for CA4+ and gadoteridol until the agents reached 40% of the cartilage depth ( Figure 4A).

| Effects of cartilage constituents to the diffusion
The correlation between the cartilage constituents and the contrast agent partitions were studied at three time-  ( Figures 5A and 6). However, based on the current results, the PGs alone does not govern the diffusion of CA4+ in the superficial and middle zones (ie, from the articulating surface to~40% of the cartilage depth). The C CA4+ max inversely correlates with cartilage water content. This might be due to the loss of PGs or an increase in the water content in the superficial and middle zones, resulting from the loss of collagen integrity. 2 As expected, the collagen concentration had no direct effect on C CA4+ max (Figures 5C and 7). Even though the depthwise gadoteridol partition resembles the water distribution in cartilage ( Figure 3B), the expected association between the water content and Thus, the extraction of diffusion coefficients will be a premise of future study requiring finite element modeling. 48 To conclude, the diffusion of cationic contrast agents depends not only on the PG concentration but also on the water content,