Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2018
Quantitative Dual Contrast CT
Technique for Evaluation of Articular Cartilage Properties
Bhattarai, A
Springer Nature
Tieteelliset aikakauslehtiartikkelit
© Biomedical Engineering Society All rights reserved
http://dx.doi.org/10.1007/s10439-018-2013-y
https://erepo.uef.fi/handle/123456789/6766
Downloaded from University of Eastern Finland's eRepository
Quantitative Dual Contrast CT Technique for Evaluation of Articular Cartilage 1
Properties 2
Abhisek Bhattarai 1, 2, Juuso T. J. Honkanen 6, Katariina A. H. Myller 1, 2, Mithilesh 3
Prakash 1, Miitu Korhonen 1, Annina E. A. Saukko 1, Tuomas Viren 6, Antti Joukainen 5, 4
Amit N. Patwa 3, 4, Heikki Kröger 5, Mark W. Grinstaff 3, Jukka S Jurvelin 1, Juha Töyräs 5
1, 2
6
1 Department of Applied Physics, University of Eastern Finland, Kuopio, Finland 7
2 Diagnostic Imaging Center, Kuopio University Hospital, Kuopio, Finland 8
3 Departments of Biomedical Engineering, Chemistry, and Medicine, Boston University, 9
Boston, MA, United States 10
4 School of Liberal Studies and Education, Navrachana University, Gujarat, India.
11
5 Department of Orthopedics, Traumatology and Hand Surgery, Kuopio University 12
Hospital, Kuopio, Finland 13
6 Cancer Center, Kuopio University Hospital, Kuopio, Finland 14
Submitted to Annals of Biomedical Engineering, December 2017 15
Corresponding author:
16
Abhisek Bhattarai, M.Sc. (Tech.) 17
Department of Applied Physics, University of Eastern Finland, Finland 18
P.O. Box 1627, 70211 Kuopio, Finland 19
Tel: +358 505922536, 20
Fax: +358 17162131 21
Email: abhisek.bhattarai@uef.fi, juuso.honkanen@uef.fi, katariina.myller@uef.fi, 22
mithilesh.prakash@uef.fi, miitu.korhonen@uef.fi, annina.saukko@uef.fi, 23
tuomas.viren@kuh.fi, antti.joukainen@kuh.fi, anpatwa@gmail.com, 24
heikki.kroger@kuh.fi, m.grin@bu.edu, jukka.jurvelin@uef.fi, juha.toyras@uef.fi 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
ABSTRACT 26
Cationic CT contrast agents detect loss of cartilage proteoglycans (PGs) more sensitively 27
than anionic or non-ionic agents. However, degeneration related loss of PGs and increase 28
in water content have opposite effects on the diffusion of the cationic agent, lowering its 29
sensitivity. In contrast to cationic agents diffusion of non-ionic agents is governed only by 30
steric hindrance and water content. We hypothesize that sensitivity of iodine-based 31
cationic agent may be enhanced by simultaneous use with non-ionic gadolinium-based 32
agent. We introduce a quantitative dual energy CT technique (QDECT) for simultaneous 33
quantification of two contrast agents in cartilage. We employ this technique to improve the 34
sensitivity of cationic CA4+ by normalizing its partition in cartilage with that of non-ionic 35
gadoteridol. The technique was evaluated with measurements of contrast agent mixtures 36
of known composition and human osteochondral samples (n=57) after immersion (72h) in 37
mixture of CA4+ and gadoteridol. Samples were arthroscopically graded and 38
biomechanically tested prior to QDECT (50kV/100kV). QDECT determined contrast agent 39
mixture compositions correlated with the true compositions (R2=0.99, average 40
error=2.27%). Normalizing CA4+ partition in cartilage with that of gadoteridol improved 41
correlation with equilibrium modulus (from ρ=0.701 to ρ=0.795). To conclude, QDECT 42
enables simultaneous quantification of iodine and gadolinium contrast agents improving 43
diagnosis of cartilage integrity and biomechanical status.
44
KEYWORDS 45
Biomechanics, Cartilage, Cationic contrast agent, Contrast enhanced computed 46
tomography, Dual energy CT 47
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
INTRODUCTION 48
Mechanical impact injury of articular cartilage (e.g. fall or sports related accidents) often 49
initiates cartilage degeneration and development of post-traumatic osteoarthritis due to 50
the limited self-repair capability of the aneural and avascular cartilage tissue.18 Using 51
present diagnostic techniques clinicians cannot sensitively detect the injury at its earliest 52
stages. Moreover, patient symptoms (pain and loss of mobility) arise, only at later stages 53
of cartilage disease progression, thus leaving clinicians with limited treatment options at 54
the time of diagnosis. Therefore, it is imperative to develop diagnostic methods capable 55
of early detection of articular cartilage injuries.
56
The first signs of cartilage injury include disruption of superficial collagen network 57
and loss of proteoglycans (PGs).1,6 Contrast enhanced computed tomography (CECT) 58
enables detection of cartilage degeneration and lesions as well as estimation of tissue 59
composition based on the diffusion of contrast agents.11,12,19 In CECT, anionic contrast 60
agents are widely used to investigate cartilage tissue properties as their partition is 61
inversely proportional to the fixed negative charge density of PGs.8,10,14 62
Recently, cationic contrast agents were introduced for CT imaging of cartilage.1,8,24 63
Cationic contrast agent molecules, such as iodine-based CA2+ and CA4+, are attracted 64
by the negative fixed charge (i.e. PGs) in cartilage, providing potentially a more sensitive 65
technique for monitoring PG content in tissue compared with use of anionic and non-ionic 66
agents. 1,15,25 However, characteristic tissue changes related to cartilage degeneration, 67
such as loss of PGs and increase of permeability, have opposite effects on the diffusion 68
of cationic agents lowering their sensitivity.
69 4
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
Along Gibbs-Donnan theory diffusion of cationic contrast agents in cartilage is 70
governed by PG content, permeability and water content, while with non-ionic agents, the 71
diffusion is only controlled by permeability and water content of tissue. To improve the 72
diagnostic potential and enhance the sensitivity of a cationic contrast agent (e.g. iodine- 73
based CA4+), it could be used simultaneously with a non-ionic agent (e.g., gadolinium- 74
based gadoteridol). By normalizing the partition of a cationic agent with that of non-ionic 75
one the diagnostic potential of the cationic agent may be maximized. Further, we 76
hypothesize that with the application of two X-ray tube voltages (50 kV and 100 kV), 77
simultaneous quantification of iodine (I) and gadolinium (Gd) based agents is possible due 78
to the element specific K-edges having significant effect on X-ray absorption.
79
The aims of the present study are to: (1) introduce a quantitative dual energy CT 80
technique (QDECT) for simultaneous determination of the partitions of iodine and 81
gadolinium-based contrast agents in a mixture; (2) determine the partitions of cationic and 82
non-ionic agents in cartilage using QDECT; (3) evaluate the change in diagnostic 83
sensitivity of the CA4+ cationic contrast agent after normalization of its partition with that 84
of a non-ionic agent; and (4) illustrate the association of partitions of contrast agents in 85
the cartilage with the tissue biomechanical properties.
86
87
MATERIAL AND METHODS 88
MicroCT scanning of Contrast Agent Solutions and Mixtures 89
MicroCT (Skyscan 1172, Skyscan, Kontich, Belgium) imaging was conducted using two 90
X-ray tube voltages (50 kV and 100 kV) and isotropic voxel size of 25 × 25 × 25 µm3. 91
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
Additional filtering was used to obtain the desired X-ray spectra. With 50 kV tube voltage, 92
0.5 mm thick aluminum (Al) and 0.05 mm thick copper (Cu) filters provided by the microCT 93
manufacturer were used along with a custom-made 0.1 mm thick Cu filter. When image 94
acquisition was conducted using 100 kV tube voltage only the 0.5 mm thick Al filter was 95
applied. Mass attenuation coefficients of contrast agents were determined at both 96
energies by imaging series of solutions with varying iodine (0, 12, 24, 48 mg/ml) and 97
gadolinium (0, 12, 24, 48 mg/ml) concentrations in distilled water (Fig. 1b).
98
Along Beer-Lambert law and Bragg’s rule of mixtures22: 99
𝐼𝐸 = 𝐼𝑜𝐸ⅇ−𝛼𝐸 , (1) 100
𝛼𝐸 = µ I𝐸 𝐶I+ µGd𝐸𝐶Gd , (2) 101
where, 𝐼𝑜𝐸 and 𝐼𝐸 are the intensities of the incident and transmitted X-ray beams through 102
material with attenuation coefficient 𝛼𝐸 at energy E, respectively. The concentrations of I 103
(𝐶I ) and Gd (𝐶Gd ) in contrast agent mixtures can be solved using the equations 3 and 4.
104
𝐶
I=
µ𝛼100µGd50−𝛼50µGd100I100µGd50−µI50µGd100
105
106 (3)
𝐶
Gd=
µ 𝛼100µI50−𝛼50µI100Gd100µI50−µGd50µI100
,
(4) 107
where, µI50 , µGd50, µ I100 and µGd100 are the mass attenuation coefficients of the iodine and 108
gadolinium-based contrast agents measured using 50 and 100 kV tube voltages.
109
The dual energy technique was first tested with measurements of contrast agent 110
mixtures of known iodine and gadolinium concentrations in distilled water [I/Gd 111
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
concentration (mg/ml) ratios of 4.8/43.2, 9.6/38.4, 14.4/33.6, 19.2/28.8, 24.0/24.0, 112
28.8/19.2, 33.6/14.4, 38.4/9.6 and 43.2/4.8]. Subsequently, the true and the solved 113
contrast agent iodine and gadolinium concentrations in the mixtures were compared (Fig.
114
1c).
115 116
Preparation and MicroCT imaging of Osteochondral Samples 117
Osteochondral samples (n = 57, d = 8 mm, Fig. 2) were drilled out from human cadaver 118
(n = 2) distal femur (n = 4) and proximal tibia (n = 4) (Decision number 150/2016, Research 119
Ethics Committee of the Northern Savo Hospital District, Kuopio University Hospital, 120
Kuopio, Finland). For the CT experiment, the samples were cut half and their edges were 121
sealed with cyanoacrylate (Loctite, Henkel Norden AB, Dusseldorf, Germany) to allow 122
contrast agent diffusion only through the articular surface. The samples were immersed 123
in an isotonic mixture of iodine-based cationic contrast agent (CA4+, q = +4, M = 1354 124
g/mol) and gadolinium-based non-ionic agent (gadoteridol, q = 0, M = 558.69 g/mol, 125
ProHanceTM, Bracco Diagnostic Inc., Monroe Twp., NJ, USA) bath (3 ml/sample) for 72 126
hours at 4 °C. The bath was continuously gently stirred during the immersion of 127
osteochondral samples. Contrast agent mixture (24 mg/ml of both iodine and gadolinium 128
in distilled water) was supplemented with inhibitors of proteolytic enzymes [5 mM 129
ethylenediaminetetraacetic acid (EDTA, VWR International, France) and penicillin- 130
streptomycin-amphotericin (Antibiotic Antimycotic solution, stabilized, Sigma-Aldrich Inc., 131
St. Louis, MO, USA)] to prevent general degradation of proteins in tissue. With the 132
established QDECT protocol human osteochondral samples were imaged with an 133
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
isotropic voxel size of 25 × 25 × 25 µm3 using a micro-CT scanner (Skyscan 1172) before 134
and after immersion in the contrast agent bath.
135 136
Determining Contrast Agent Distribution 137
Micro-CT images were reconstructed using NRecon software (Bruker co., Kontich, 138
Belgium). Further microCT image data analysis was conducted using MATLAB (R2016b, 139
Mathworks Inc., Natick, MA, USA). Cartilage was segmented from subchondral bone 140
manually using segmentation software (Seg3D2 vs 2.2, The University of Utah, Salt Lake 141
City, UT, USA) and custom made MATLAB code to select the cartilage surface and 142
cartilage-bone interface 23. Volume of interest (2000 µm × 1250 µm × cartilage thickness) 143
was selected to obtain the depthwise X-ray attenuation profile. For each sample, native 144
cartilage X-ray attenuation profile was subtracted from that measured after the contrast 145
agent diffusion. The concentrations of iodine (𝐶I ) and gadolinium (𝐶Gd ) in cartilage were 146
determined using the equations 3 and 4, respectively.
147 148
Biomechanical testing and ICRS grading 149
A custom made material testing device equipped with an actuator having resolution of 0.1 150
µm (PM500-1 A, Newport, Irvine, CA, USA) and a load cell with resolution of 0.005 N 151
(Sensotec, Columbus, OH, USA) was employed for biomechanical testing of 152
osteochondral samples. During testing the samples were immersed in phosphate buffered 153
saline supplemented with inhibitors of proteolytic enzymes [5 mM EDTA, VWR 154
International and 5 mM benzamidine hydrochloride hydrate (Sigma-Aldrich Inc.)]. A flat 155
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
ended metallic indenter [d = 728 µm (n = 29) or d = 667 µm (n = 28)] was driven in 156
perpendicular contact with the articular surface using a pre stress of 12.5 kPa 13. Stress 157
relaxation protocol consisting of four compressive steps (each being 5% of cartilage 158
thickness, 100%/s ramp rate) was implemented with a 900 s relaxation after each step.
159
The equilibrium modulus (Eequilibrium, fit to the equilibrium points of the last three steps) and 160
instantaneous modulus (Einstantaneous, the ramp phase of the third step) were calculated 161
using the Hayes model7. Following stress strain relaxation tests, dynamic moduli (Edynamic) 162
was determined based on Hayes model using sinusoidal loading (f = 1 Hz, strain 163
amplitude = 2 % of cartilage thickness). An experienced orthopaedic surgeon (A.J.) 164
graded all the cartilage sample locations before sample extraction using the ICRS 165
(International Cartilage Repair Society) grading (scale 0 to 4) 3. 166
167
Statistical Analysis 168
Pearson’s correlation was used to analyze the relation between the true and measured 169
iodine and gadolinium concentrations in contrast agent mixtures. Spearman’s rho (ρ) was 170
determined to analyze the relation between contrast agent partitions in cartilage and the 171
values of biomechanical moduli. All the statistical analysis were conducted using SPSS 172
(vs. 23, SPSS Inc., IBM Company, Armonk, NY, USA).
173
174
RESULTS 175
Composition of the contrast agent mixture, as determined with the dual energy technique, 176
correlated linearly with the true mixture composition (R2 = 0.99, P < 0.01), with an average 177
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
error of 2.27 percentage points (Fig. 1c). After immersing human osteochondral samples 178
(n = 57) for 72 hours in the contrast agent mixture, the partitions (mean ± SD) of CA4+
179
and gadoteridol in cartilage were 197.6 ± 28.4% and 66.0 ± 9.2%, respectively (Table 1).
180
CA4+ and gadoteridol partitions determined with the dual energy technique showed 181
increasing and decreasing trends, respectively, along cartilage depth (Fig. 3). CA4+ and 182
gadoteridol partitions in the cartilage superficial zone (500 µm) correlated significantly 183
(0.656 ≤ ρ ≤ 0.701, P < 0.001) and (-0.566 ≥ ρ ≥ -0.583, P < 0.001), respectively, with the 184
cartilage biomechanical properties (Table 2). Importantly, normalizing CA4+ partition with 185
that of gadoteridol in the superficial cartilage improved linear correlations with the 186
biomechanical parameters, e.g. with Eequilibrium from ρ = 0.701 to ρ = 0.795 (Table 2).
187
Furthermore, normalizing CA4+ partition in full thickness cartilage with that of gadoteridol 188
improved the correlation with ICRS grading (from ρ = -0.385 to ρ = -0.458) (Table 2).
189 190
DISCUSSION 191
In this laboratory study, a QDECT methodology for simultaneous quantification of iodine 192
(CA4+) and gadolinium (gadoteridol) contrast agents in human articular cartilage was 193
developed and validated. We hypothesized that this novel technique would enable 194
simultaneous determination of the depthwise distribution of the two contrast agents.
195
Furthermore, we summarized that the diagnostic sensitivity of the cationic contrast agent 196
would improve by normalizing its partition with that of the non-ionic agent. Our 197
experimental findings confirmed both hypotheses.
198
Cartilage degeneration in osteoarthritis (OA) involves loss of proteoglycans (PGs), 199
disruption in the collagen network, and increase in cartilage water content. 5 In OA, loss 200
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
of PGs and increase in water content occur simultaneously having opposite effects on the 201
diffusion of cationic agents. Unfortunately, this limits the diagnostic value of cationic 202
contrast agents at clinically relevant imaging time points (<1 hour after contrast agent 203
administration, i.e. when complete diffusion has not taken place).
204
In the present experiment, the equilibrium partition of the cationic agent (CA4+) 205
was nearly three times that of the non-ionic agent (gadoteridol) and increased towards the 206
deep cartilage. This is in line with literature reports, as cationic contrast agents have 207
shown an uptake in direct proportion to the negative fixed charge density (i.e. PG content), 208
which are in greater concentrations in middle and deep zones of cartilage.17,20,26 On the 209
contrary, the partition of non-ionic gadoteridol decreased as a function of cartilage depth.
210
This was also expected as cartilage water content decreases with increasing cartilage 211
tissue depth.2,4 The permeability of cartilage to electrically neutral agents reduces with 212
increasing cartilage tissue depth because of the increasing steric hindrance of contrast 213
agent molecules induced by the collagen network and the highly concentrated PGs.16,18,21 214
Hence, the quantification of non-ionic agent’s partition along the tissue depth provides 215
information of water content and permeability of the cartilage extracellular matrix.
216
Simultaneous determination of the partitions of both agents with the dual energy 217
technique enabled normalization of cationic agent partition with that of the non-ionic agent.
218
In line with our hypothesis, the normalization of the CA4+ partition in superficial cartilage 219
improved its correlation with cartilage biomechanical properties (Eequilibrium from ρ = 0.701 220
to ρ = 0.795, Edynamic modulus from ρ = 0.656 to ρ = 0.748) and structural integrity (ICRS 221
grade from ρ = -0.398 to ρ = -0.408). Even though the trend towards higher correlations 222
after normalization was observed with most reference parameters, the enhancement in 223
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
correlation was not statistically significant. This is possibly due to the relatively low sample 224
number (n = 57) in the present study. Correlation between biomechanical moduli and 225
normalized CA4+ in full thickness cartilage did not demonstrate enhancement in CA4+
226
sensitivity as with the cartilage superficial layer. This could be due to superficial layer 227
controlling cartilage indentation response.9,13 228
Based on these results, the introduced QDECT technique may have significant 229
potential for clinical diagnostics of degenerative cartilage conditions. However, the current 230
study protocol does possess limitations. The dual energy CT imaging was conducted at 231
diffusion equilibrium (osteochondral plugs immersed for 72 hours in contrast agent bath).
232
In normal clinical practice, contrast-enhanced CT scanning is carried out in earlier time 233
points, i.e. at 45 minutes after administration of the contrast agent.10,19 During the first 234
hour, the diffusion is fast and the depthwise partitions of the contrast agents are changing 235
rapidly.8 In such a scenario, imaging with two energies must be done as instantaneously 236
and simultaneously as possible. This is not a problem with clinical scanners, but might 237
jeopardize the reliability of the results when using high resolution microCT-scanners with 238
longer imaging times. In the present study, imaging was conducted at diffusion equilibrium 239
and therefore this is not a problem. However, this issue warrants further research and the 240
potential of the present technique should be tested in all phases of the diffusion process 241
and most importantly in clinically relevant time points.
242
To conclude, QDECT imaging enables simultaneous determination of the 243
distribution of iodine and gadolinium-based contrast agents in cartilage. Importantly, the 244
introduced technique improves the diagnostic sensitivity of contrast-enhanced imaging of 245
cartilage.
246 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
ACKNOWLEDGEMENTS 247
Sandra Sefa, (B.Sc.) is acknowledged for assistance with the biomechanical 248
measurements. Jaakko Sarin, M.Sc. (Tech) is acknowledged for assistance in sample 249
extraction. Academy of Finland (Projects 269315, 307932), Kuopio University Hospital 250
(VTR 5041746, 5041757, PY210), Instrumentarium Science Foundation (170033) and 251
Doctoral Program in Science, Technology and Computing (SCITECO, University of 252
Eastern Finland) are acknowledged for financial support.
253 254
CONFLICTS OF INTERESTS 255
The authors have no conflicts of interest.
256
257
REFERENCES 258
1. Bansal, P. N., R. C. Stewart, V. Entezari, B. D. Snyder, and M. W. Grinstaff.
259
Contrast agent electrostatic attraction rather than repulsion to glycosaminoglycans 260
affords a greater contrast uptake ratio and improved quantitative CT imaging in 261
cartilage. Osteoarthr. Cartil. 19:970–976, 2011.
262
2. Berberat, J. E., M. J. Nissi, J. S. Jurvelin, and M. T. Nieminen. Assessment of 263
interstitial water content of articular cartilage with T1 relaxation. Magn. Reson.
264
Imaging 27:727–732, 2009.
265
3. Brittberg, M., and C. S. Winalski. Evaluation of cartilage injuries and repair. J.
266
Bone Joint Surg. Am. 85–A Suppl:58–69, 2003.
267 268 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
4. Brocklehurst, R., M. T. Bayliss, A. Maroudas, H. L. Coysh, M. A. Freeman, P. A.
269
Revell, and S. Y. Ali. The composition of normal and osteoarthritic articular 270
cartilage from human knee joints. With special reference to unicompartmental 271
replacement and osteotomy of the knee. J. Bone Joint Surg. Am. 66:95–106, 272
1984.
273
5. Buckwalter, J. A., H. J. Mankin, and A. J. Grodzinsky. Articular cartilage and 274
osteoarthritis. Instr. Course Lect. 54:465–480, 2005.
275
6. Ewers, B. J., V. M. Jayaraman, R. F. Banglmaier, and R. C. Haut. Rate of blunt 276
impact loading affects changes in retropatellar cartilage and underlying bone in 277
the rabbit patella. J. Biomech. 35:747–755, 2002.
278
7. Hayes, W. C., L. M. Keer, G. Herrmann, and L. F. Mockros. A mathematical 279
analysis for indentation tests of articular cartilage. J. Biomech. 5:541–551, 1972.
280
8. Honkanen, J. T. J., M. J. Turunen, J. D. Freedman, S. Saarakkala, M. W.
281
Grinstaff, J. H. Ylarinne, J. S. Jurvelin, and J. Töyräs. Cationic Contrast Agent 282
Diffusion Differs Between Cartilage and Meniscus. Ann. Biomed. Eng. 44:2913–
283
2921, 2016.
284
9. Julkunen, P., R. K. Korhonen, W. Herzog, and J. S. Jurvelin. Uncertainties in 285
indentation testing of articular cartilage: a fibril-reinforced poroviscoelastic study.
286
Med. Eng. Phys. 30:506–515, 2008.
287
10. Kokkonen, H. T., H. C. Chin, J. Töyräs, J. S. Jurvelin, and T. M. Quinn. Solute 288
Transport of Negatively Charged Contrast Agents Across Articular Surface of 289
Injured Cartilage. Ann. Biomed. Eng. , 2016.doi:10.1007/s10439-016-1756-6 290
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
4 291 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
292
11. Kokkonen, H. T., J. S. Jurvelin, V. Tiitu, and J. Töyräs. Detection of mechanical 293
injury of articular cartilage using contrast enhanced computed tomography.
294
Osteoarthr. Cartil. 19:295–301, 2011.
295
12. Kokkonen, H. T., J.-S. Suomalainen, A. Joukainen, H. Kroger, J. Sirola, J. S.
296
Jurvelin, J. Salo, and J. Töyräs. In vivo diagnostics of human knee cartilage 297
lesions using delayed CBCT arthrography. J. Orthop. Res. 32:403–412, 2014.
298
13. Korhonen, R. K., M. S. Laasanen, J. Töyräs, R. Lappalainen, H. J. Helminen, and 299
J. S. Jurvelin. Fibril reinforced poroelastic model predicts specifically mechanical 300
behavior of normal, proteoglycan depleted and collagen degraded articular 301
cartilage. J. Biomech. 36:1373–1379, 2003.
302
14. Kulmala, K. A. M., H. M. Karjalainen, H. T. Kokkonen, V. Tiitu, V. Kovanen, M. J.
303
Lammi, J. S. Jurvelin, R. K. Korhonen, and J. Töyräs. Diffusion of ionic and non- 304
ionic contrast agents in articular cartilage with increased cross-linking—
305
Contribution of steric and electrostatic effects. Med. Eng. Phys. 35:1415–1420, 306
2013.
307
15. Lakin, B. A., H. Patel, C. Holland, J. D. Freedman, J. S. Shelofsky, B. D. Snyder, 308
K. S. Stok, and M. W. Grinstaff. Contrast-enhanced CT using a cationic contrast 309
agent enables non-destructive assessment of the biochemical and biomechanical 310
properties of mouse tibial plateau cartilage. J. Orthop. Res. 34:1130–1138, 2016.
311
16. Maroudas, A. Biophysical chemistry of cartilaginous tissues with special reference 312
to solute and fluid transport. Biorheology 12:233–248, 1975.
313 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
4 314 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
17. Maroudas, A., H. Muir, and J. Wingham. The correlation of fixed negative charge 315
with glycosaminoglycan content of human articular cartilage. Biochim. Biophys.
316
Acta 177:492–500, 1969.
317
18. Muir, H., P. Bullough, and A. Maroudas. The distribution of collagen in human 318
articular cartilage with some of its physiological implications. J. Bone Joint Surg.
319
Br. 52:554–563, 1970.
320
19. Myller, K. A. H., M. J. Turunen, J. T. J. Honkanen, S. P. Vaananen, J. T. Iivarinen, 321
J. Salo, J. S. Jurvelin, and J. Töyräs. In Vivo Contrast-Enhanced Cone Beam CT 322
Provides Quantitative Information on Articular Cartilage and Subchondral Bone.
323
Ann. Biomed. Eng. 45:811–818, 2017.
324
20. Palmer, A. W., R. E. Guldberg, and M. E. Levenston. Analysis of cartilage matrix 325
fixed charge density and three-dimensional morphology via contrast-enhanced 326
microcomputed tomography. Proc. Natl. Acad. Sci. U. S. A. 103:19255–19260, 327
2006.
328
21. Pouran, B., V. Arbabi, A. A. Zadpoor, and H. Weinans. Isolated effects of external 329
bath osmolality, solute concentration, and electrical charge on solute transport 330
across articular cartilage. Med. Eng. Phys. 38:1399–1407, 2016.
331
22. Rangacharyulu, C. Physics of Nuclear Radiations Concepts, Techniques and 332
Applications. Taylor and Francis, 2013, 130 pp.
333 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
23. Saukko, A. E. A., J. T. J. Honkanen, W. Xu, S. P. Vaananen, J. S. Jurvelin, V.-P.
334
Lehto, and J. Töyräs. Dual Contrast CT Method Enables Diagnostics of Cartilage 335
Injuries and Degeneration Using a Single CT Image. Ann. Biomed. Eng. , 336
2017.doi:10.1007/s10439-017-1916-3.
337
24. Stewart, R. C., P. N. Bansal, V. Entezari, H. Lusic, R. M. Nazarian, B. D. Snyder, 338
and M. W. Grinstaff. Contrast-enhanced CT with a high-affinity cationic contrast 339
agent for imaging ex vivo bovine, intact ex vivo rabbit, and in vivo rabbit cartilage.
340
Radiology 266:141–150, 2013.
341
25. Stewart, R. C., J. T. J. Honkanen, H. T. Kokkonen, V. Tiitu, S. Saarakkala, A.
342
Joukainen, B. D. Snyder, J. S. Jurvelin, M. W. Grinstaff, J. Toyras, and J. Töyräs.
343
Contrast-Enhanced Computed Tomography Enables Quantitative Evaluation of 344
Tissue Properties at Intrajoint Regions in Cadaveric Knee Cartilage. Cartilage 345
8:391–399, 2017.
346
26. Stewart, R. C., A. N. Patwa, H. Lusic, J. D. Freedman, M. Wathier, B. D. Snyder, 347
A. Guermazi, and M. W. Grinstaff. Synthesis and Preclinical Characterization of a 348
Cationic Iodinated Imaging Contrast Agent (CA4+) and Its Use for Quantitative 349
Computed Tomography of Ex Vivo Human Hip Cartilage. J. Med. Chem. 60:5543–
350
5555, 2017.
351 352 353 354 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
Table 1. Values of biomechanical moduli (instantaneous, equilibrium and dynamic) and contrast 355
agent partitions (mean±SD) in superficial and full thickness cartilage in osteochondral samples 356
extracted from tibial and femoral locations in human cadaver (n = 2) knee joints (n = 4).
357
FLC (Femoral Lateral Condyle), FMC (Femoral Medial Condyle), FG (Femoral Groove), TLC (Tibial Lateral 358
Condyle), TMC (Tibial Medial Condyle).
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379
Modulus (MPa) Contrast Agent Partition (%)
Location Thickness
(mm) Instantaneous Equilibrium Dynamic 1Hz
Superficial Cartilage (top 500 µm)
Full Thickness Cartilage
CA4+ Gadoteridol CA4+ Gadoteridol FLC (n = 9) 2.86±0.61 18.76±7.10 1.46±0.92 8.51±3.45 157.7±26.6 78.4±5.1 212.7±17.9 61.6±4.6 FMC (n = 9) 2.63±0.59 16.64±9.82 1.51±1.31 7.65±4.87 154.2±39.7 81.5±6.4 212.2±33.1 68.1±9.8 FG (n = 9) 2.61±0.41 13.59±9.56 1.12±0.92 6.81±4.69 129.1±22.2 77.6±6.2 209.3±17.9 62.3±9.7 TMC (n = 14) 2.48±0.42 12.06±9.30 0.69±0.59 4.85±3.63 125.8±25.9 84.1±9.8 201.1±23.5 69.4±9.8 TLC (n = 16) 2.65±0.84 10.50±10.55 0.65±0.62 4.42±4.08 111.4±20.6 82.4±10.9 171.1±22.2 66.5±9.3 Total (n = 57) 2.63±0.61 13.64±9.64 1.02±0.94 6.10±4.43 131.8±31.4 81.3±8.7 197.6±28.4 66.0±9.2 4
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
Table 2. Linear correlations (Spearman’s rho) between the equilibrium, instantaneous and dynamic 380
moduli and contrast agent partition in cartilage (n = 57).
381
382
* P < 0.05 383
384
** P < 0.01 385 386
387 388 389 390 391 392 393 394 395 396 397 398 399
Superficial Cartilage (Top 500 µm) Full Thickness Cartilage CA4+
Normalized
CA4+ Gadoteridol CA4+
Normalized
CA4+ Gadoteridol
Equilibrium Modulus (MPa) 0.795** 0.701** -0.566** 0.364** 0.492** -0.182 Instantaneous Modulus (MPa) 0.693** 0.614** -0.573** 0.317* 0.359** -0.175 Dynamic Modulus 1Hz (MPa) 0.748** 0.656** -0.583** 0.381** 0.428** -0.208
ICRS Grading -0.408** -0.398** 0.249 -0.458** -0.385** 0.317*
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
400
Figure 1. (a) The simulated microCT spectra (SPEKTR vs. 3, I-star Lab, John Hopkins University, 401
Baltimore, MD, USA) at tube voltages of 50 kV and 100 kV are presented along mass attenuation 402
curves for Iodine (I) and Gadolinium (Gd). (b) Linear fits between contrast agents (I and Gd) 403
concentrations and X-ray attenuation at 50 and 100 kV tube voltages are α I (50 kV) = 0.0528 CI + 404
0.1094, α Gd (50 kV) = 0.0432 CGd + 0.0621, α I (100 kV) = 0.0353 CI + 0.0985 and α Gd (100 kV) = 0.0508 405
CGd + 0.071. (c) Linear correlation between the true mixture compositions and the mixture 406
compositions determined using the QDECT technique.
407
408 409 410 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
411 412
Figure 2. Osteochondral plugs (d = 8 mm) were extracted from the marked locations of human 413
cadaver (n = 2) knee joints (n = 4).
414
415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
430
Figure 3. (a) Depthwise partitioning of (a) gadoteridol and (b) CA4+ in cartilage after 72 hours 431
of immersion in mixture of cationic and non-ionic contrast agents. The black solid lines 432
represent the mean partition in all the samples (n = 57).
433
434 435 436 437
438 439 440 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
441
442 443 444 445 446
Figure 4. Correlations (Spearman’s rho) between CA4+ partition (normalized and non-normalized by
gadoteridol partition) in cartilage and cartilage biomechanical properties. The partition values are calculated for superficial cartilage (i.e. top 500 µm from cartilage surface). Linear fit is drawn to enable better visualization of correlation. FG (Femoral Groove), FLC (Femoral Lateral Condyle), FMC (Femoral Medial Condyle), TLC (Tibial Lateral Condyle), TMC (Tibial Medial Condyle).
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
