Accepted Manuscript Title: Magnetic resonance compositional imaging of articular cartilage: what can we expect in veterinary medicine? Author: Fanny Hontoir, Peter Clegg, Jean-Francois Nisolle, Simon Tew, JeanMichel Vandeweerd PII: DOI: Reference:

S1090-0233(15)00177-X http://dx.doi.org/doi:10.1016/j.tvjl.2015.04.035 YTVJL 4504

To appear in:

The Veterinary Journal

Accepted date:

28-4-2015

Please cite this article as: Fanny Hontoir, Peter Clegg, Jean-Francois Nisolle, Simon Tew, JeanMichel Vandeweerd, Magnetic resonance compositional imaging of articular cartilage: what can we expect in veterinary medicine?, The Veterinary Journal (2015), http://dx.doi.org/doi:10.1016/j.tvjl.2015.04.035. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Review Magnetic resonance compositional imaging of articular cartilage: What can we expect in veterinary medicine? Fanny Hontoir a, Peter Clegg b, Jean-Francois Nisolle c, Simon Tew b, Jean-Michel Vandeweerd a, * a

Integrated Veterinary Research Unit (IRVU), Department of Veterinary Medicine, Faculty of Sciences, University of Namur, Rue de Bruxelles 61, 5000 Namur, Belgium b Dept. of Musculoskeletal Biology, Faculty of Health and Life Sciences, Leahurst Campus, University of Liverpool, Neston, UK c Université Catholique de Louvain, Hôpital Mont-Godinne, Yvoir, Belgium

* Corresponding author. Tel.: +32 81 72 43 79. E-mail address: [email protected] (J-M. Vandeweerd).

22 23 24

Highlights 

25 26

composition of cartilage. 

27 28



33

dGEMRIC and T2 mapping seem to be most applicable for compositional imaging of animal cartilage.



31 32

Technical issues still limit the use of some techniques for veterinary research or clinical practice.

29 30

Compositional magnetic resonance imaging of cartilage aims to determine biochemical

GAG-CEST and sodium imaging are better used with high field magnets, which have limited availability.



Long acquisition times are sometimes required, for instance in T1ρ and diffusion-weighted imaging.

34

Page 1 of 31

35 36

Abstract Since cartilage has limited ability to repair itself, it is useful to determine its

37

biochemical composition early in clinical cases. It is also important to assess cartilage content

38

in research animals in longitudinal studies in vivo. In recent years, compositional imaging

39

techniques using magnetic resonance imaging (MRI) have been developed to assess the

40

biochemical composition of cartilage. This article describes MR compositional imaging

41

techniques, and discusses their use and interpretation.

42 43

Technical concerns still limit the use of some techniques for research and clinical use,

44

especially in veterinary medicine. Glycosaminoglycan chemical-exchange saturation transfer

45

and sodium imaging are better used with high field magnets, which have limited availability.

46

Long acquisition times are sometimes required, for instance in T1rho (ρ) and diffusion-

47

weighted imaging, and necessitate general anaesthesia. Even in human medicine, some

48

techniques such as ultra-short echo T2 are not fully validated, and nearly all techniques

49

require validation for veterinary research and clinical practice. Delayed gadolinium-enhanced

50

MRI of cartilage and T2 mapping appear to be the most applicable methods for compositional

51

imaging of animal cartilage. Combining T2 mapping and T1ρ allows for the assessment of

52

proteoglycans and the collagen network, respectively.

53 54

Keywords: Cartilage; Collagen; Imaging; MRI; Proteoglycans

55

Page 2 of 31

56

Introduction

57

Osteoarthritis (OA) is a degenerative process of the joint characterised by progressive

58

degeneration of the articular cartilage and reduced joint function. Articular cartilage has

59

biomechanical properties that are attributable to its extracellular matrix (ECM), which is

60

composed of collagen, proteoglycans (PGs), hyaluronan (HA) and water (Lu and Mow,

61

2008).

62 63

Collagen fibres are organised into three zones (Fig. 1). In the radial zone, the fibres are

64

perpendicular to the articular surface, forming rows; in the transitional zone, they overlap as

65

thin lamellae, and in the superficial zone, they are tangential to the articular surface. This

66

arcade-like orientation is involved in the dynamic behaviour of cartilage, reducing stresses in

67

its deep part during loading and ensuring resistance to shearing forces in its superficial part

68

(Halonen et al., 2013).

69 70

PGs are composed of a central core protein linked to hundreds of negatively charged

71

polysaccharide chains called glycosaminoglycans (GAGs) and to HA (Fig. 1). Sulfate and

72

carboxyl groups of GAGs carry negative charges (Maroudas et al., 1969). Each fixed negative

73

charge requires a positive counter-ion (Na+) for the tissue to maintain overall

74

electroneutrality. The high concentration of Na+ results in attraction of water.

75 76

Water linked to Na+ is called free water. It exists in two other forms when it is directly

77

linked to collagen or PGs (Maroudas, 1976). When the joint is loaded, water flows through

78

the ECM. Its flow is limited by the frictional resistance of collagen and the attraction by ions.

79

Water contributes toward the viscoelastic behaviour of cartilage (Lu and Mow, 2008).

80

Page 3 of 31

81

With progression of OA, synthesis is insufficient to compensate for the degradation of

82

ECM; biochemical changes, such as a decrease in PGs and the degradation of collagen, occur

83

as a consequence (Bijlsma et al., 2011). The alteration of the collagen network and the

84

associated increase in water content modify the biomechanical properties of cartilage.

85 86

Since cartilage has limited ability for repair, biochemical changes associated with

87

early OA should be identified as soon as possible in clinical cases. For research purposes, it

88

would be extremely useful to determine the composition of cartilage in longitudinal studies in

89

vivo. In recent years, compositional imaging techniques using magnetic resonance (MR)

90

imaging have been developed to assess the biochemical composition of cartilage. This article

91

reviews MR compositional imaging techniques and discusses their use and possible

92

applications to veterinary medicine.

93 94

T2 mapping

95

The mobility of water protons varies with tissue type; it is high when protons are in

96

free water and low when they are immobilised in ECM. This influences the transverse (T2)

97

relaxation time, due to spin-spin (neighbouring) interactions (Watrin et al., 2001). In MR

98

imaging (MRI) sequences highlighting T2 (T2 weighted, W), mobile water protons (e.g. in

99

synovial fluid or in damaged collagen networks with increased free water content) give a

100

hypersignal (long T2), whereas water protons immobilised in ECM (short T2) give a

101

hyposignal (David-Vaudey et al., 2004). Since T2 reflects water content and integrity,

102

specialised T2 sequences have been created in an attempt to identify the early stages of OA.

103

In general terms, T2 sequences can demonstrate pathology because they identify changes in

104

water content.

105

Page 4 of 31

106

Signals can be plotted in a T2 decay curve, which is used to calculate a

107

mathematical parameter called the T2 relaxation time of the tissue. T2 mapping involves

108

multiple T2 sequences with slightly varying parameters. The differences in T2 relaxation time

109

over all sequences are then combined and a colour map of T2 times is generated. The

110

application of T2 mapping to cartilage assessment was initially reported in 2000 (Mosher et

111

al., 2000) and can be performed with a 1.5 Tesla system.

112 113

T2 is highly sensitive to changes in hydration and it is able detect a temporary

114

physiological loss of water after exercise (Liess et al., 2002; Mosher et al., 2010; Stehling et

115

al., 2010). T2 values are higher in the radial zone of elderly people due to higher water

116

mobility, secondary to changes in the structure of PG aggregates (Mosher et al., 2000).

117 118

T2 values are also influenced by the orientation of collagen fibrils. A lack of signal

119

occurs when fibres are so well aligned that the amount of signal produced in a T2 weighted

120

image is changed, a phenomenon called isotropy. This is very evident with a 9.4 T magnet,

121

but is less apparent in clinical systems (Nissi et al., 2006). For example, imaging the same

122

osteochondral plug placed at angles of 3°, 25°, 40° and 57° with respect to the z-axis produces

123

T2 values which vary from 28 ms to 56 ms. In the radial zone, where collagen fibrils are

124

highly aligned and perpendicular to the articular surface, T2 varies up to 80% with the

125

direction of the main magnetic field. In the transitional zone, where collagen fibrils are less

126

aligned, T2 is less influenced by plug orientation (Xia, 1998). This orientational effect seems

127

to be less pronounced in vivo; it has been suggested that compression of cartilage during

128

normal daily activity results in modifications to the orientation of the collagen fibres in the

129

radial zone (Mosher et al., 2001). These differences should be taken into account when

130

comparing data from in vitro and in vivo studies.

Page 5 of 31

131 132

Several studies have demonstrated the utility of T2 mapping for the assessment of

133

osteoarthritic cartilage. In human cadaver knee specimens, T2 values correlated to

134

macroscopic, histological and mechanical assessments and increased with OA grade (David

135

Vaudey et al., 2004; Lammentausta et al., 2007). The orientational effect was used in clinical

136

research to assess the structural organisation of cartilage before and after treatment. T2

137

mapping has also been used to monitor cartilage repair over time in patients, especially to

138

assess the development of a collagen network with zonal organisation similar to normal

139

cartilage (Smith 2001; Kelly et al., 2006; Trattnig et al., 2007a; Welsch et al., 2008a;

140

Domayer et al., 2010).

141 142

In veterinary medicine, several studies have investigated possible applications of T2

143

mapping. The accuracy of the technique to measure cartilage thickness of the distal

144

metacarpus/metatarsus was determined in equine cadaver limbs (Carstens et al., 2013a). The

145

authors concluded that T2 mapping was accurate for measuring thickness, except at locations

146

where the cartilage was in direct contact with the cartilage of the proximal phalanx. This

147

exception was explained by the fact that, when windowing images to identify the articular

148

surface, partial volume averaging of the thin intra-articular space (often on the 1-pixel

149

threshold) tended to determine the boundary between the articular surfaces to be halfway

150

between them, resulting in a halving of the measured distance. It was concluded that future

151

research could be conducted to assess whether higher resolution images could be obtained

152

with a 3.0 Tesla (T) system, which could result in superior measurement of the thin (0.8-0.9

153

mm) cartilage in that anatomical region.

154

Page 6 of 31

155

Another study determined the intra- and inter-observer variability of T2 mapping to

156

assess the articular cartilage in the region of the medial coronoid process (MCP) of six healthy

157

dogs (Wucherer et al., 2012). The histology and PG content of MCP were used to confirm

158

normal articular cartilage, but not to evaluate the accuracy of the technique to assess the

159

biochemical composition of articular cartilage. The study demonstrated that T2 mapping was

160

reproducible (i.e. it had low variability) and also established a normal reference range of T2

161

values (56 ± 8 ms) for healthy cartilage of the MCP.

162 163

The utility of the technique at 1.5 T to evaluate cartilage repair has been investigated

164

in several clinical studies in horses. It helped differentiate deep, middle and superficial

165

cartilage in reparative fibrous tissue sites (mean T2 values 53.3, 58.6 and 54.9 ms,

166

respectively) from normal hyaline cartilage (40.7, 53.6 and 61.6 ms, respectively) after

167

arthroscopic osteochondral autograft transplantation and microfracture arthroplasty (White et

168

al., 2006). The effects of those procedures were also assessed in a pony osteochondral model.

169

While there were increasing T2 values from the deep to the superficial layer of normal hyaline

170

cartilage had (range, 40.7-61.6 ms), as expected, fibrous tissue from cartilage repair did not

171

show the same depth-dependent variation in T2 values. There was a good agreement between

172

T2 pattern (presence or absence of a depth-dependent organisation) and histological findings

173

(Clark et al., 2010). The same group also used T2 mapping data as an outcome measure in a

174

study evaluating direct delayed human adenoviral bone morphogenic protein (BMP)-2 or

175

BMP-6 gene therapy for bone and cartilage regeneration in the pony model (Menendez et al.,

176

2011). T2 mapping of the defect produced high T2 values, corresponding with the

177

disorganisation of collagen repair. Another group showed that T2 values were longer in the

178

centre of cartilage repair in an equine model where an osteochondral defect was created in the

179

lateral trochlear ridge (Pownder et al., 2011).

Page 7 of 31

180 181 182

T2* mapping In T2* mapping, then the radiofrequency (RF) pulse is 90°, the longitudinal

183

magnetisation disappears and transverse magnetisation appears. The curve that represents the

184

time needed for the transverse magnetisation to decrease should represent T2. This would be

185

the case if the interaction between protons (spin-spin interaction) were the only reason for the

186

reduction in phase coherence and for increasing entropy. However, the microscopic

187

heterogeneity of the main field (i.e. constant heterogeneity) accelerates dephasing. Therefore,

188

the signal decreases at a faster time (called T2*). This is why a 180° pulse must follow the

189

90° pulse (in a T2W sequence) if the true T2 characteristics of the tissue are to be measured

190

(McRobbie et al., 2006; Hoa et al., 2010).

191 192

One advantage of T2* mapping is the shorter acquisition time compared to T2

193

mapping. In a comparison study where field of view, matrix, slice thickness, and voxel size

194

were identical, acquisition time for T2 mapping was 5 min 11 s vs. 2 min 35 s for T2*

195

mapping (Welsch et al., 2011). It is important to note that T2* is the inverse of the T2 signal,

196

so the information gained from T2* images might not be the same as from T2 images, since

197

the inhomogeneity of the signal is measured more than the true T2 signal. T2* has been tested

198

in vivo in human knee (Hughes et al., 2007; Welsch et al., 2008b; Mamisch et al., 2012) and

199

hip joints (Bittersohl et al., 2009). This technique has not yet been assessed in veterinary

200

medicine for articular cartilage.

201 202 203 204

Ultra-short echo (UTE) - T2 mapping Deep and calcified zones of articular cartilage, with their highly ordered collagen fibrils, limit the motion of protons. Protons are therefore more influenced by spin-spin

Page 8 of 31

205

interactions and lose phase coherence more quickly. T2 is very short, between 1 ms and 2 ms

206

(Du et al., 2009). While standard T2 mapping uses echo times of 10 ms or more, UTE-T2

207

mapping uses echo times shorter than 1 ms (Bae et al., 2010; Williams et al., 2010). It has the

208

potential to visualise cartilage at the osteochondral junction, focusing on deep cartilage and

209

calcified cartilage. UTE-T2 values are most likely more sensitive to matrix disorganisation

210

than to matrix content, as they did not seem to correlate with cartilage composition (collagen

211

or GAG) in one study (Williams et al., 2010). UTE-T2 mapping has been performed at 3

212

Tesla with an acquisition time of 4-10 min (Gao et al., 2013). The technique has not been

213

used in veterinary medicine and although it could have research applications for visualising

214

the cartilage-bone interface, its clinical utility needs further investigation.

215 216

T1rho (ρ) mapping In the T1ρ technique, a supplementary low amplitude RF pulse is applied for a

217 218

prolonged period of time during relaxation and appropriately aligned with the transverse

219

magnetisation (Charagundla, 2004). The pulse, called spin-locking pulse, alters the tendency

220

of the spin components to process at their own individual frequencies, instead locking them in

221

an orientation aligned with the spin-locking pulse (Fig. 2). The magnetisation no longer

222

relaxes according to T2, but instead relaxes according to a time called T1ρ. The prevention of

223

dephasing slows down magnetisation decay dramatically, resulting in a T1ρ that is longer than

224

T2.

225 226

In the T1ρ technique, due to the spin-locking pulse, a new small magnetic

227

environment is created where energy transfer can occur. In the presence of the spin-locking

228

pulse, transverse magnetisation can exchange energy with lattice (magnetic environment)

229

processes occurring at the spin-locking frequency, which is directly proportional to the

Page 9 of 31

230

amplitude of the spin-locking pulse. Since there are far more processes (movements) in the

231

lattice operating at the low spin-locking frequencies (so-called ‘slow motions’) than at the

232

Larmor frequency, more energy can be lost and T1ρ is shorter than T1 (Charagundla, 2004).

233 234

Because slow motions in the lattice are typically associated with large molecules, T1ρ

235

is sensitive to the protein composition of tissue, and could therefore provide information

236

about the macromolecular properties of tissue such as GAGs content. Some authors have

237

suggested that T1ρ and T2 imaging might be complementary since T1ρ seems to be a more

238

specific indicator of PG content, while T2 is more sensitive to collagen network modification

239

(Taylor et al., 2008). The technique can be performed with a 1.5 system (Duvvuri et al, 2001).

240

In a study using a 3.0 Tesla MRI unit (Stahl et al., 2009), the acquisition time was

241

approximately 13 min.

242 243

In a human cadaveric study, T1ρ moderately and variably correlated with the content

244

of the GAG and r² ranged from 0.13 to 0.84, depending on the specimen (Keenan et al.,

245

2011). In live subjects, T1ρ values were significantly higher in patients with OA compared to

246

those that were healthy (Li et al., 2005; Regatte et al., 2006; Stahl et al., 2009).

247 248

When T1ρ imaging was tested in a live porcine model of OA, following the

249

injection of interleukin-1β, T1ρ seemed to correlate with the induced PG loss. This study

250

highlighted difficulties encountered in reaching adequate resolution while preserving an

251

acceptable acquisition time (Wheaton et al., 2004). Other authors have also stressed the

252

problem of sequence length. Due to the long-duration of the spin-locking pulse, T1ρ is

253

sensitive to inhomogeneities (called B1 inhomogeneities; Charagundla, 2004). Moreover, the

254

long duration of the RF can reach the allowable specific absorption rate (SAR) of RF energy

Page 10 of 31

255

(Wheaton et al., 2004). The human SAR limits, in Europe, range from 0.08 W/kg (whole-

256

body) to 20.0 W/kg (limb extremities) for an average duration of 6 min. SAR limits are made

257

to limit the amount of energy absorbed by the human body and this limit can only be

258

extrapolated to animals.

259 260

In veterinary research, T1ρ has been used in a preclinical model of equine cartilage

261

repair. Cadaver stifle specimens showed higher T1ρ values in the centre of the cartilage repair

262

area relative to normal cartilage, both in the superficial zone (66 ± 22 ms vs. 42 ± 5 ms) and

263

in the deep zone (57 ± 19 ms vs. 33 ± 4 ms; Pownder et al., 2011). T1ρ values were only

264

slightly reduced at the interface. The authors suggested this could reflect the non-uniformity

265

of the tissue in the centre of the repair and incorporation with the native tissue at the periphery

266

of the repair. It was acknowledged this hypothesis needed to be confirmed by histological

267

analysis.

268 269 270

Diffusion-weighted imaging (DWI) DWI is based on the varying motion properties of water protons in different

271

environments. Protons can move in all three directions in free water, but they are restricted by

272

collagen or PG barriers in cartilage (Hagman et al., 2006). In DWI, a RF pulse is applied with

273

decreasing intensity with distance, the so-called ‘gradient pulse’. If the structure of the tissue

274

allows diffusion (Fig. 3), such as when the direction of the gradient is parallel to the collagen

275

fibres, the gradient pulse induces motion of the water protons in the direction of the gradient

276

pulse. Water protons also gain energy from the RF pulse in proportion to the distance.

277 278 279

A 180° RF pulse (spin-echo) is then applied, followed by a second gradient pulse with the same decreasing intensity as the first. Since all water protons do not move the same

Page 11 of 31

280

distance due to the different intensities of pulse, the second gradient pulse is not able to bring

281

them back to their initial position. The MRI signal in the new position (and energy state) is

282

different of that of the initial position; the system measures this difference.

283 284

If diffusion is not allowed by the structure of the tissue (Fig. 3), such as when the

285

direction of the gradient is perpendicular to the collagen fibres, motion by water protons is not

286

allowed. After the 180° RF pulse and second gradient pulse, water protons are found in their

287

initial position at the same level of energy. In this new setting, no loss of signal is detected.

288 289

The setting where diffusion is allowed in both directions is observed when, for

290

example, collagen fibres have disappeared. The combination of three different directions of

291

gradient is necessary to thoroughly capture the mobility of water and indirectly provide an

292

image of the 3D integrity of the tissue network (Bammer, 2003).

293 294

The DWI pulsed-gradient spin-echo sequence is characterised by several parameters:

295

the length (duration) of the gradient pulse, the time-period between both gradient pulses, and

296

the amplitude of the gradient pulse (Bammer, 2003). The three parameters are summarised by

297

a mathematical coefficient called the diffusion-sensitising factor (b-value), which is used to

298

calculate the apparent diffusion coefficient (ADC; Xia et al., 1995). A low ADC is associated

299

with the restriction of water molecules by the environment, e.g. by a healthy intact collagen

300

network in cartilage (Burstein et al., 1993). This sequence can be performed with a 1.5 Tesla

301

system. In one study that used a 3.0 Tesla unit, acquisition time was approximately 6-7 min

302

(Apprich et al., 2012).

303

Page 12 of 31

304

ADC has been shown to correlate positively with OA in human cartilage samples

305

(Mlynarik et al., 2003). DWI has also been used in vivo for follow up assessment of surgical

306

cartilage repair in human patients (Mamisch et al., 2008; Friedrich et al., 2010). Interestingly,

307

a canine study has been reported (Xia et al., 1995). Four millimetre discs of cartilage were

308

harvested from both humeral heads of an 8-month old, skeletally mature, disease-free

309

Labrador retriever, and from an osteoarthritic 4-year old Labrador retriever. The observation

310

was made that ADC was approximately 25% higher in osteoarthritic cartilage than in the

311

surrounding cartilage. This increase was mimicked by enzymatic degradation of normal

312

cartilage with trypsin, hyaluronidase, and collagenase, or by mechanical means, suggesting

313

that ADC is more specific for the assessment of cartilage matrix integrity than selective GAG

314

loss or collagen loss.

315 316

Delayed gadolinium-enhanced MR imaging of cartilage (dGEMRIC) Gadopentetate dimeglumine (Gd-DTPA2-), a contrast agent, can shorten T1 relaxation;

317 318

an increased signal was demonstrated in a study comparing pre- and post-contrast images

319

after IV injection of the agent (Bashir et al., 1996). Since the contrast is negatively charged

320

and its concentration varies inversely to the concentration of the negatively charged GAGs,

321

T1 is shorter in GAG-depleted cartilage. T1 values are mapped and translated in a colour

322

scale. It can also be administered intra-articularly, but requires a large dilution (Kwack et al.,

323

2008).

324 325

The principal disadvantage of dGEMRIC is the delay needed for Gd-DTPA2- to

326

penetrate through the cartilage (Gold et al., 2009). For example, a 90 min delay is required for

327

the human knee cartilage (Welsch et al., 2008b; Domayer et al., 2010). The Gd-DTPA2-

328

penetration is also influenced by mobilisation of the joint and it is increased after exercise

Page 13 of 31

329

(Tiderius et al., 2004). Penetration is also better in thinner cartilage (Hawezi et al., 2011). The

330

technique has already been used for a number of years. In vitro research has demonstrated that

331

T1 values in dGEMRIC correlate well with GAG concentration in human cartilage (Bashir et

332

al., 1996; Miyata et al., 2006) and that dGEMRIC predicted changes in Young’s modulus in

333

plugs of bovine humeral head (Nieminen et al., 2004) and in human knee plugs (Kurkijarvi et

334

al., 2004). Additionally, longitudinal clinical studies have used dGEMRIC for ongoing

335

assessment of autologous chondrocytes transplants in humans (Gillis et al., 2001; Watanabe et

336

al., 2006; Trattnig et al., 2007b; Wiewiorski et al., 2013).

337 338

In veterinary cartilage research, dGEMRIC was first used to assess the effects of

339

administration of human adenoviral (Ad) BMP-2 or BMP-6 in experimentally created femoral

340

condyle lesions in ponies (Menendez et al., 2011). The authors demonstrated that dGEMRIC

341

T1 relaxation times were greater in lesions treated with Ad-BMP6 than in saline treated

342

controls. Reference values for dGEMRIC T1 values have been obtained for normal articular

343

cartilage near the medial coronoid process in dogs with a 3.0 T magnet (400 ± 47 ms;

344

Wucherer et al., 2012). In that study, dogs received IV Gd-DTPA2- under general anaesthesia

345

and passive motion of the joint was performed for 10 min. The time between IV injection and

346

the start of the dGEMRIC sequence ranged from 47-82 min.

347 348

Another group assessed the dGEMRIC characteristics of the cartilage of the distal

349

metacarpus/metatarsus after intra-articular injection of contrast in 20 cadaver specimens

350

collected from normal Thoroughbred horses (Carstens et al., 2013a). Significant differences

351

were observed between anatomical sites and this was attributed to differences in cartilage

352

substrate that could have occurred as an adaptation to the varying forces applied. Since

353

dGEMRIC T1 values at the sites with significant differences decreased to a minimum at 60-

Page 14 of 31

354

120 min post-contrast, this was proposed as the optimal time for post-contrast scans. T1

355

values were measured between 642-674 ms after 60 min post-contrast, which is longer than

356

the 400 ms used for the canine MCP (Wucherer et al., 2012). This discrepancy could be due

357

to the method of injection, cartilage thickness or other inherent cartilage differences between

358

the two species and joints. There was also poor spatial resolution of the thin distal

359

metacarpus/metatarsus. The authors suggested that dedicated surface coil, decreased field of

360

view, or a higher field strength magnet could be used to improve resolution.

361 362

The same group also used dGEMRIC to determine the cartilage thickness of the distal

363

metacarpus/metatarsus in equine cadaver limbs with a 1.5 T system (Carstens et al., 2013b).

364

The conclusions were similar as those for T1ρ mapping: the technique was accurate only at

365

locations where the cartilage of the distal metacarpus/metatarsus is not in direct contact with

366

the proximal phalanx cartilage.

367 368

Sodium imaging

369

In this technique, MRI targets the sodium nucleus, which is comprised of 11 protons

370

and 12 neutrons; the unequal number of protons and neutrons confers magnetic properties to

371

sodium. In cartilage, sodium content is proportional to PG content, since it counter-balances

372

the negative fixed charge density of GAGs. A loss of PG induces a decrease in FCD and thus

373

a loss of sodium (Shapiro et al., 2002).

374 375

Though a better signal is obtained with high field magnets, imaging can be performed

376

with a 1.5 T system (Hani et al., 2013). Because the spin frequency of the sodium nucleus

377

differs from that of hydrogen, a dedicated system is required to reach resonance (Krusche-

378

Mandl et al., 2012; Madelin et al., 2012). The system is used mostly for research and uses

Page 15 of 31

379

long acquisition times of approximately 20-30 min (Krusche-Mandl et al., 2012; Newbould et

380

al., 2013).

381 382

Sodium MRI was able to detect chemically-induced PG loss in animal articular

383

cartilage (Borthakur et al., 2000; Wheaton et al., 2004). It has been tested in vivo at 7T in the

384

human knee (Wang et al., 2009) and it has been used to follow up matrix-associated

385

autologous chondrocyte transplantation in human knee (Trattnig et al., 2010).

386 387 388

GAG chemical-exchange saturation transfer (GAG-CEST) sequence In conventional MRI, signals are detected from the protons of components (such as

389

water) that have sufficiently long T2 relaxation times (> 10 ms). The T2 of the restricted

390

protons (with low mobility), such as protons of the amide group (-NH) and hydroxyl groups (-

391

OH) of GAGs, are too short (< 1 ms) to be detected.

392 393

An effect called ‘chemical exchange saturation transfer’ (CEST) refers to the induced

394

exchange of protons between water and protons of specific molecules, such as GAGs (Fig. 1).

395

For the transfer to happen, a selective non-resonant RF pulse must be applied (Fig. 4). The

396

protons of low energy in GAGs are excited in such a way that populations of the higher and

397

lower energy levels become equal. When this occurs, further absorption of energy is not

398

possible and there is no net magnetization; the spin system is termed ‘saturated’. Saturated

399

protons have the property to exchange with unsaturated protons of water. The selective

400

saturation of exchangeable protons requires high field strength (3.0 or 7T), otherwise

401

saturation of GAG molecules and water molecules would occur at the same time and no

402

exchange would be possible (Zhou and van Zijl, 2006; Borthakur and Reddy, 2010). This

403

exchange (CEST effect) causes a change in magnetisation of water and a signal difference

Page 16 of 31

404

that can be recorded. This technique is also affected by pH changes, since H+ and OH-

405

contribute to the overall chemical exchange. It is also temperature-dependent (Ward et al.,

406

2000) and acquisition time is around 15 min (Krusche-Mandl et al., 2012). GAG-CEST has

407

been tested for human knee imaging (Schmitt et al., 2011; Krusche-Mandl et al., 2012; Singh

408

et al., 2012).

409 410 411

Applications in veterinary medicine The current review demonstrates that compositional imaging of cartilage in animals is

412

primarily undertaken for research purposes, using small or large animal models. Technical

413

concerns still limit the use of some techniques both for research and clinical use. Gag-CEST

414

and sodium imaging necessitate high field magnets and long acquisition times (approximately

415

15 min and 20 min, respectively). The UTE-T2 mapping has not been investigated in

416

veterinary research. DWI was initially tested in vitro on fragments of canine cartilage in 1995

417

(Xia et al., 1995), but has not received much attention since. It appears that T2 mapping, T1ρ

418

and dGEMRIC are the most applicable to clinical veterinary medicine.

419 420

T2 mapping can be performed at 1.5 T with an acceptable acquisition time (about 4

421

min) and is easily applicable in veterinary medicine. The intra- and inter-observer variability

422

of T2 mapping is low (Wucherer et al., 2012) and its ability to differentiate repair cartilage

423

from normal cartilage has been demonstrated in animal studies (White et al., 2006; Clark et

424

al., 2010; Menendez et al., 2011; Pownder et al., 2011). However, human studies have shown

425

that several factors can influence T2 values, such as age (Mosher et al., 2000), exercise level

426

(Mosher et al., 2010; Stehling et al., 2010) and the orientation of the collagen fibrils to the Z-

427

axis (Nissi et al., 2006) and therefore the anatomy and the curvature of the articular surface. In

428

addition, this orientation effect can differ in vivo and ex vivo (Mosher et al., 2001). Validation

Page 17 of 31

429

of the technique for each species and joint to account for the aforementioned variables is

430

required. Moreover, identical study protocols must be used to enable comparison. The use of

431

T2* mapping could shorten acquisition time, but it remains to be determined if information

432

from T2* images is similar to that obtained with T2 images. In addition, correlations with the

433

biochemical composition of the cartilage have not been demonstrated in animals, so it is

434

possible that T2 times are more reflective of the network and the state of water than the

435

concentration in water or PG.

436 437

T1ρ is potentially complementary to T2 mapping, as it is more correlated to PG

438

content, while T2 is an indicator of collagen network. Human T1ρ studies have demonstrated

439

the challenge of reaching adequate resolution in a reasonable acquisition time despite the use

440

of a 3.0 T magnet (14 min; Charagundla et al., 2004, Wheaton et al., 2004). Its use in

441

veterinary medicine has been reported in just one study in cadaver specimens (Pownder et al.,

442

2011) and the sequence is not available for all systems (e.g. Siemens).

443 444

A 1.5 T system has been used with dGEMRIC, but the resolution was suboptimal. In

445

addition, Gd-DTPA2- penetration varies with injection technique (IV or intra-articular), time

446

after administration, mobilisation of the joint (Tiderius et al., 2004), thickness of cartilage

447

(Hawezi et al., 2011), or other inherent cartilage differences between species and joints

448

(Carstens et al., 2013a, b). This should be taken into account when studies are compared.

449 450

The poor spatial resolution of the thin cartilage in some animal joints is a frequent

451

difficulty. The cartilage of the metatarso/metacarpophalangeal joint in horses, for example, is

452

0.8-0.9 mm (Carstens et al., 2013 b). This technical problem occurs not only with the

453

compositional MRI techniques discussed in the current review, but also with conventional

Page 18 of 31

454

MRI, even when using high field magnets and sequences that are considered optimal for the

455

study of cartilage (Hontoir et al., 2014). Cartilage thickness can also vary with anatomical

456

site. In a study of 12 7-year old purebred Merino sheep, the mean articular cartilage thickness

457

was 0.43 mm in the abaxial part of the lateral tibial condyle, 1.10 mm in the axial part of the

458

lateral tibial condyle, 1.17 mm in the axial part of the medial tibial condyle and 0.56 mm in

459

the abaxial part of the medial tibial condyle (Appleyard et al., 2003). Cartilage thickness also

460

varies with species. In one study, cartilage was 2.2 mm thick in human medial femoral

461

condyles, 2.0 mm in horses and 0.5 mm in sheep (Frisbie et al., 2006). Another study in

462

crossed Texel sheep reported that the thickness of the femoral condyle cartilage varied from

463

1.5 mm (medial condyle) to 1.0 mm (lateral condyle; Pirson et al., 2014). These studies

464

indicate that animal cartilage thicknesses can be very different from those in humans,

465

resulting in poor inter-species generalisability of techniques.

466 467

Higher field strength magnets can improve resolution since more protons are available

468

and the signal is reinforced, enabling smaller pixels. It is also possible to increase the size of

469

the matrix, but this involves increasing acquisition time. Zero filling interpolation can also be

470

used, which is the substitution of zeroes for unmeasured data points to increase the matrix size

471

of new data prior to Fourier transformation of MR data. This results in pixels smaller than the

472

actual resolution of the image. To illustrate the feasibility of T2 mapping and dGEMRIC, we

473

provide images of the ovine knee obtained by our group with a 3.0T in crossed Texel

474

cadavers. Interpolation was used and resolution was 0.14  0.14  2 mm for T2 mapping (TR

475

900; TE 16.9, 33.8, 50.7, 67.6, 84.5, 101.4; averages 4, bandwidth 237, acquisition time 18

476

min; Fig. 5) and 0.2  0.2  1.5 mm for dGEMRIC (T1 vibe 3D, TR 15, TE 3.21, flip angle 5,

477

averages 3, bandwidth 289, acquisition time 20 min) after IV injection of 15 mL gadolinium

Page 19 of 31

478

(0.5 mmol/mL) 120 min before imaging (Fig. 6). Those examples show that it is possible to

479

reach high resolution, but acquisition time remains long, which is challenging in live animals.

480 481 482

Conclusions So far, T2 mapping and dGEMRIC seem to be most applicable for compositional

483

imaging of animal cartilage. Since the thickness of the cartilage varies with species, joints and

484

anatomic sites in animals can be small and resolution must be optimised. The combination of

485

T2 mapping and T1ρ might allow for the assessment of both PGs and the collagen network,

486

respectively. Compositional imaging is currently a research tool rather than a clinical tool and

487

it usually requires a high field magnet. Confirmation that it can detect the early stages of OA

488

will be an important step in the justification of its clinical use. Compositional imaging has

489

been tested in joints used as models, but its application should be improved for joints with

490

clinical interest (e.g. the equine metacarpophalangeal joint). Because of the relative thinness

491

of cartilage in specific joints, increased availability of powerful imaging systems is required if

492

compositional imaging is to be made more applicable to practice.

493 494 495 496

Conflict of interest statement None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.

497 498 499 500

Acknowledgements This study was supported by the University of Namur (UNamur) and NARILIS (Namur Research Institute for Life Science).

501 502

References

Page 20 of 31

503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551

Appleyard, R.C., Burkhardt, D., Ghosh, P., Read, R., Cake, M., Swain, M.V., Murrell, G.A., 2003. Topographical analysis of the structural, biochemical and dynamic biomechanical properties of cartilage in an ovine model of osteoarthritis. Osteoarthritis and Cartilage 11, 65-77. Apprich, S., Trattnig, S., Welsch, G.H., Noebauer-Huhmann, I.M., Sokolowski, M., Hirschfeld, C., Stelzeneder, D., Domayer, S., 2012. Assessment of articular cartilage repair tissue after matrix-associated autologous chondrocyte transplantation or the microfracture technique in the ankle joint using diffusion-weighted imaging at 3Tesla. Osteoarthritis and Cartilage 20, 703-711. Bae, W.C., Dwek, J.R., Znamirowski, R., Statum, S.M., Hermida, J.C., D’Lima, D.D., Sah, R.L., Du, J., Chung, C.B., 2010. Ultrashort echo time MR imaging of osteochondral junction of the knee at 3 T: identification of anatomic structures contributing to signal intensity. Radiology 254, 837–845. Bammer, R., 2003. Basic principles of diffusion-weighted imaging. European Journal of Radiology 45, 169-184. Bashir, A., Gray, M. L., Burstein, D., 1996. Gd-DTPA2- as a measure of cartilage degradation. Magnetic Resonance in Medicine 36, 665-673. Bijlsma, J.W.J., Berenbaum, F., Lafeber, F.P.J.G., 2011. Osteoarthritis: an update with relevance for clinical practice. The Lancet 377, 2115-2126. Bittersohl, B., Hosalkar, H.S., Hughes, T., Kim, Y.-J., Werlen, S., Siebenrock, K.A., Mamisch, T.C., 2009. Feasibility of T2* mapping for the evaluation of hip joint cartilage at 1.5 T using a three-dimensional gradient-echo sequence: a prospective study. Magnetic Resonance in Medicine 62, 896-901. Borthakur, A., Shapiro, E.M., Beers, J., Kudchodkar, S., Kneeland, J.B., Reddy, R., 2000. Sensitivity of MRI to proteoglycan depletion in cartilage: comparison of sodium and proton MRI. Osteoarthritis and Cartilage 8, 288-293. Borthakur, A., Reddy, R., 2010. Imaging cartilage physiology. Topics in Magnetic Resonance imaging 21, 291-296. Burstein, D., Gray, M.L., Hartman, A.L., Gipe, R., Foy, B.D., 1993. Diffusion of small solutes in cartilage as measured by nuclear magnetic resonance (NMR) spectroscopy and imaging. Journal of Orthopaedic Research 11, 465-478. Carstens, A., Kirberger, R.M., Velleman, M., Dahlberg, L.E., Fletcher, L., Lammentausta, E., 2013a. Feasibility for mapping cartilage t1 relaxation times in the distal metacarpus3/metatarsus3 of thoroughbred racehorses using delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC): normal cadaver study. Veterinary Radiology and Ultrasound 54, 365–372. Carstens, A., Kirberger, R.M., Dahlberg, L.E., Prozesky, L., Fletcher, L., Lammentausta, E., 2013b. Validation of delayed gadolinium-enhanced imaging of cartilage and T2 mapping

Page 21 of 31

552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599

for quantifying distal metacarpus/metatarsus cartilage thickness in thoroughbred racehorses. Veterinary Radiology and Ultrasound 54, 139–148. Charagundla, S.R., 2004. T1ρ weighted magnetic resonance imaging: Principles and diagnostic application. Applied Radiology Suppl. 33, 32-43. Clark, D.J., Jia, G., Menendez, M.I., Choi, S., Miller, C.J., Sammet, S., Flanigan, D.C., Bertone, A.L., Knopp, M.V., 2010. In vivo quantification of cartilage regeneration in an equine model at 3T following gene therapy. Proceedings of the International Society for Magnetic Resonance in Medicine 18, 232. David-Vaudey, E., Ghosh, S., Ries, M., Majumdar, S., 2004. T2 relaxation time measurements in osteoarthritis. Magnetic Resonance Imaging 22, 673-682. Duvvuri, U., Charagundla, S.R., Kudchodkar, S.B., Kaufman, J.H., Kneeland, J.B., Rizi, R., Leigh, J.S., Reddy, R., 2001. Human knee: in vivo T1(rho)-weighted MR imaging at 1.5 T - preliminary experience. Radiology 220, 822-826. Domayer, S.E., Welsch, G.H., Nehrer, S., Chiari, C., Dorotka, R., Szomolanyi, P., Mamisch, T.C., Yayon, A., Trattnig, S., 2010. T2 mapping and dGEMRIC after autologous chondrocyte implantation with a fibrin-based scaffold in the knee: preliminary results. European Journal of Radiology 73, 636-642. Du, J., Takahashi, A.M., Chung, C.B., 2009. Ultrashort TE spectroscopic imaging (UTESI): application to the imaging of short T2 relaxation tissues in the musculoskeletal system. Journal of Magnetic Resonance Imaging 29, 412-421. Friedrich, K.M., Mamisch, T.C., Plank, C., Langs, G., Marlovits, S., Salomonowitz, E., Trattnig, S., Welsch, G., 2010. Diffusion weighted imaging for the follow up of patients after matrix associated autologous chondrocyte transplantation. European Journal of Radiology 73, 622–628. Frisbie, D.D., Cross, M.W., McIlwraith, C.W., 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. Veterinary and Comparative Orthopaedics and Traumatology 19, 142-146. Gao, S., Du, J., Wang, F., Bao, S., 2013. Magnetic resonance ultrashort echo time spin-echo imaging of the deepest layers of articular cartilage. Science China. Life Sciences 56, 672674. Gillis, A., Bashir, A., McKeon, B., Scheller, A., Gray, M.L., Burstein, D., 2001. Magnetic resonance imaging of relative glycosaminoglycan distribution in patients with autologous chondrocyte transplants. Investigative Radiology 36, 743-748. Gold, G.E., Chen, C.A., Koo, S., Hargreaves, B.A., Bangerter, N.K., 2009. Recent advances in MRI of articular cartilage. American Journal of Roentgenology 193, 628-638.

Page 22 of 31

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649

Hagmann, P., Jonasson, L., Maeder, P., Thiran, J.-P., Wedeen, V.J., Meuli, R., 2006. Understanding diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond. Radiographics 26, S205-S223. Halonen, K.S., Mononen, M.E., Jurvelin, J.S., Töyräs, J., Korhonen, R.K., 2013. Importance of depth-wise distribution of collagen and proteoglycans in articular cartilage – A 3D finite element study of stresses and strains in human knee joint. Journal of Biomechanics 46, 1184-1192. Hani, A.F., Kumar, D., Malik, A.S., Razak, R., 2013. Physiological assessment of in vivo human knee articular cartilage using sodium MR imaging at 1.5 T. Magnetic Resonance Imaging 31, 1059-1067. Hoa, D., Micheau, A., Gahide, G., Le Bars, E., Taourel, P., 2008. Chapitre 1 – Résonance magnétique nucléaire (RMN), IN : l’IRM pas à pas, Sauramps Medical, Montpellier, France, pp. 3-10. Hontoir, F., Nisolle, J.F., Meurisse, H., Simon, V., Tallier, M., Vanderstricht, R., Antoine, N., Piret, J., Clegg, P., Vandeweerd, J.M., 2014. A comparison of 3-T magnetic resonance imaging and computed tomography arthrography to identify structural cartilage defects of the fetlock joint in the horse. The Veterinary Journal 199, 115-122. Hughes, T., Welsch, G.H., Trattnig, S., Brandi, L., Domayer, S., Mamisch, T.C., 2007. T2star relaxation as a means to differentiate cartilage repair tissue after microfracturing therapy. Proceedings of the International Society for Magnetic Resonance in Medicine 15, 812. Keenan, K. E., Besier, T. F., Pauly, J. M., Han, E., Rosenberg, J., Smith, R. L., Delp, S. L., Beaupre, G. S., Gold, G. E., 2011. Prediction of glycosaminoglycan content in human cartilage by age, T1 rho and T2 MRI. Osteoarthritis and Cartilage 19, 171-179. Kelly, B. T., Potter, H. G., Deng, X.-H., Pearle, A. D., Turner, A. S., Warren, R. F., Rodeo, S. A., 2006. Meniscal allograft transplantation in the sheep knee. The American Journal of Sports Medicine 34, 1464-1477. Krusche-Mandl, I., Schmitt, B., Zak, L., Apprich, S., Aldrian, S., Juras, V., Friedrich, K.M., Marlovits, S., Weber, M., Trattnig, S., 2012. Long-term results 8 years after autologous osteochondral transplantation: 7 T gagCEST and sodium magnetic resonance imaging with morphological and clinical correlation. Osteoarthritis and Cartilage 20, 357-363. Kurkijarvi, J.E., Nissi, M.J., Kiviranta, I., Jurvelin, J.S., Nieminen, M.T., 2004. Delayed Gadolinium-enhanced MRI of cartilage (dGEMRIC) and T2 characteristics of human knee articular cartilage: Topographical variation and relationships to mechanical properties. Magnetic Resonance in Medicine 52, 41-46. Kwack, K.S., Cho, J.H., Kim, M., Yoon, C.S., Yoon, Y.S., Choi, J.W., Kwon, J.W., Min, B.H., Sun, J.S., Kim, S.Y., 2008. Comparison study of intraarticular and intravenous gadolinium-enhanced magnetic resonance imaging of cartilage in a canine model. Acta Radiologica 49:65-74.

Page 23 of 31

650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697

Lammentausta, E., Kiviranta, P., Töyräs, J., Hyttinen, M.M., Kiviranta, I., Nieminen, M.T., Jurvelin, J.S., 2007. Quantitative MRI of parallel changes of articular cartilage and underlying trabecular bone in degeneration. Osteoarthritis and Cartilage 15, 1149-1157. Li, X., Han, E.T., Ma, C.B., Link, T.M., Newitt, D.C., Majumdar, S., 2005. In vivo 3T spiral image based multi-slice T1ρ mapping of knee cartilage in osteoarthritis. Magnetic Resonance in Medicine 54, 929-936. Liess, C., Lüsse, S., Karger, N., Heller, M., Glüer, C.-C., 2002. Detection of changes in cartilage water content using MRI T2-mapping in vivo. Osteoarthritis and Cartilage 10, 907-913. Lu, X.L., Mow V.C., 2008. Biomechanics of articular cartilage and determination of material properties. Medicine and Science in Sports and Exercise 40, 193-199. Madelin, G., Babb, J.S., Xia, D., Chang, G., Jerschow, A., Regatte, R.R., 2012. Reproducibility and repeatability of quantitative sodium magnetic resonance imaging in vivo in articular cartilage at 3T and 7T. Magnetic Resonance in Medicine 68, 841-849. Mamisch, T.C., Menzel, M.I., Welsch, G.H., Bittersohl, B., Salomonowitz, E., Szomolanyi, P., Kordelle, J., Marlovits, S., Trattnig, S., 2008. Steady-state diffusion imaging for MR in-vivo evaluation of reparative cartilage after matrix-associated autologous chondrocyte transplantation at 3 tesla—preliminary results. European Journal of Radiology 65, 72–79. Mamisch, T.C., Hughes, T., Mosche, T.J., Mueller, C., Trattnig, S., Boesch, C., Welsch, G.H., 2012. T2star relaxation times for assessment of articular cartilage at 3T: a feasibility study. Skeletal Radiology 41, 287-292. Maroudas, A., Muir, H., Wingham, J., 1969. The correlation of fixed negative charge with glycosaminoglycan content of human articular cartilage. Biochimica and Biophysica Acta 177, 492-500. Maroudas, A.I., 1976. Balance between swelling pressure and collagen tension in normal and degenerate cartilage. Nature 260, 808-809. McRobbie, D.W., Moore, E.A., Graves, M.J., Prince, M.R., 2006. Chapter 8 – Getting in tune : resonance and relaxation, IN : MRI : from picture to proton (2nd edition), Cambridge University Press, New York, USA, pp. 137-166. Menendez, M.I., Clark, D.J., Carlton, M., Flanigan, D.C., Jia, G., Sammet, S., Weisbrode, S.E., Knopp, M.V., Bertone, A.L., 2011. Delayed human adenoviral BMP-2 or BMP-6 gene therapy for bone and cartilage regeneration in a pony osteochondral model. Osteoarthritis and Cartilage 19, 1066-1075. Miyata, S., Homma, K., Numano, T., Furukawa, K., Tateishi, T., Ushida, T., 2006. Assessment of fixed charge density in regenerated cartilage by Gd-DTPA-enhanced MRI. Magnetic Resonance in Medical Sciences 5, 73-78.

Page 24 of 31

698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747

Mlynarik, V., Sulzbacher, I., Bittsansky, M., Fuiko, R., Trattnig, S., 2003. Investigation of apparent diffusion constant as an indicator of early degenerative disease in articular cartilage. Journal of Magnetic Resonance Imaging 17, 440-444. Mosher, T.J., Dardzinski, B.J., Smith, M.B., 2000. Human articular cartilage: influence of aging and early symptomatic degeneration on the spatial variation of T2- Preliminary findings at 3 T. Radiology 214, 259-266. Moscher, T.J., Smith, H., Dardzinski, B.J., Schmithorst, V.J., Smith, M.B., 2001. MR imaging and T2 mapping of femoral cartilage: in vivo determination of the magical angle effect. American Journal of Roentgenology 177, 665-669. Mosher, T.J., Liu, Y., Torok, C.M., 2010. Functional cartilage MRI T2 mapping: Evaluating the effect of age and training on knee cartilage response to running. Osteoarthritis and Cartilage 18, 358-364. Nieminen, M.T., Töyräs, J., Laasanen, M.S., Silvennoinen, J., Helminen, H.J., Jurvelin, J.S., 2004. Prediction of biomechanical properties of articular cartilage with quantitative magnetic resonance imaging. Journal of Biomechanics 37, 321-328. Newbould, R.D., Miller, S.R., Upadhyay, N., Rao, A.W., Swann, P., Gold, G.E., Strachan, R.K., Matthews, P.M., Taylor, P.C., Brown, A.P., 2013. T1-weighted sodium MRI of the articulator cartilage in osteoarthritis: a cross sectional and longitudinal study. PLoS One 8, e73067. Nissi, M.J., Rieppo, J., Töyräs, J., Laasanen, M.S., Kiviranta, I., Jurvelin, J.S., Nieminen, M.T., 2006. T2 relaxation time mapping reveals age- and species-related diversity of collagen network architecture in articular cartilage. Osteoarthritis and Cartilage 14, 12651271. Pirson, R., Matagne, A., Nisolle, J.F., Hontoir, F., Meirlaen, P., Gustin, P., Kirschvink, N., Clegg, P., Vandeweerd, J.M., 2014. What does multiplication of tidemarks mean in the ovine femoro-tibial joint? Proceedings of the WCO (World Congress on Osteoarthritis and Osteoporosis), Sevilla, Spain, p 172. Pownder, S., Koff, M.F., Fortier, L., Castiglione, E., Saska, R., Bradica, G., Novakofski, K., Potter, H.G., 2011. Quantitative and morphologic evaluation of cartilage repair in an equine model. Proceedings of the International Society for Magnetic Resonance in Medicine 19, 499. Regatte, R.R., Akella, S.V., Lonner, J.H., Kneeland, J.B., Reddy, R., 2006. T1rho relaxation mapping in human osteoarthritis (OA) cartilage: comparison of T1rho with T2. Journal of Magnetic Resonance Imaging 23, 547-553. Schmitt, B., Zbyn, S., Stelzeneder, D., Jellu, V., Paul, D., Lauer, L., Bachert, P., Trattnig, S., 2011. Cartilage quality assessment by using glycosaminoglycan chemical exchange saturation transfer and 23Na MR imaging at 7T. Radiology 260, 257-264. Shapiro, E.M., Borthakur, A., Gougoutas, A.J., Reddy, R., 2002. 23Na MRI accurately measures FCD in articular cartilage. Magnetic Resonance in Medicine 47, 284-291.

Page 25 of 31

748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797

Singh, A., Haris, M., Cai, K., Kassey, V.B., Kogan, F., Reddy, D., Hariharan, H., Reddy, R., 2012. Chemical exchange saturation transfer magnetic resonance imaging of human knee cartilage at 3 T and 7 T. Magnetic Resonance in Medicine 68, 588-594. Smith, H.E., Moscher, T.J., Dardzinski, B.J., Collins, B.G., Collins, C.M., Yang, Q.X., Schmithorst, V.J., Smith, M.B., 2001. Spatial variation in cartilage T2 of the knee. Journal of Magnetic Resonance Imaging 14, 50-55. Stahl, R., Luke, A., Li, X., Carballido-Gamio, J., Ma, C.B., Majumdar, S., Link, T.M., 2009. T1rho, T2 and focal knee cartilage abnormalities in physically active and sedentary healthy subjects versus early OA patients - A 3.0-Tesla MRI study. European Radiology 19, 132-143. Stehling, C., Liebl, H., Krug, R., Lane, N.E., Nevitt, M.C., Lynch, J., McCulloch, C.E., Link TM., 2010. Patellar cartilage: T2 values and morphologic abnormalities at 3.0-T MR imaging in relation to physical activity in asymptomatic subjects from the osteoarthritis initiative. Radiology 254, 509-520. Taylor, C., Carballido-Gamio, J., Majumdar, S., Li, X., 2008. Comparison of quantitative imaging of cartilage for osteoarthritis: T2, T1ρ, dGEMRIC and contrast-enhanced computed tomography. Magnetic Resonance Imaging 27, 779-784. Tiderius, C.J., Svensson, J., Leander, P., Ola, T., Dahlberg, L., 2004. dGEMRIC (Delayed Gadolinium-Enhanced MRI of Cartilage) indicates adaptive capacity of human knee cartilage. Magnetic Resonance in Medicine 51, 286-290. Trattnig, S., Mamisch, T.C., Welsch, G., Glaser, C., Szomolanyi, P., Gebetsroither, S., Stastny, O., Horger, W., Millington, S., Marlovits, S., 2007a. Quantitative T2 mapping of matrix-associated autologous chondrocyte transplantation at 3T: an in vivo crosssectional study. Investigative Radiology 42, 442-448. Trattnig, S., Marlovits, S., Gebetsroither, S., Szomolanyi, P., Welsch, G.H., Salomonowitz, E., Watanabe, A., Deimling, M., Mamisch, T.C., 2007b. Three-dimensional delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) for in vivo evaluation of reparative cartilage after matrix-associated autologous chondrocyte transplantation at 3.0T: Preliminary results. Journal of Magnetic Resonance Imaging 26, 974-982. Trattnig, S., Welsch, G.H., Juras, V., Szomolanyi, P., Mayerhoefer, M.E., Stelzeneder, D., Mamisch, T.C., Bieri, O., Scheffler, K., Zbyn, S., 2010. 23 Na MR imaging at 7T after knee matrix-associated autologous chondrocyte transplantation: Preliminary results. Radiology 257, 175-184. Wang, L., Wu, Y., Chang, G., Oesingmann, N., Schweitzer, M.E., Jerschow, A., Regatte, R.R., 2009. Rapid isotropic 3D-sodium MRI of the knee joint in vivo at 7T. Journal of Magnetic Resonance Imaging 30, 606-614. Ward, K.M., Aletras, A.H., Balaban, R.S., 2000. A new class of contrast agents for MRI based on proton chemical exchange dependent saturation transfer (CEST). Journal of Magnetic Resonance 143, 79-87.

Page 26 of 31

798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847

Watanabe, A., Wada, Y., Obata, T., Ueda, T., Tamura, M., Ikehira, H., Moriya, H., 2006. Delayed gadolinium-enhanced MR to determine glycosaminoglycan concentration in reparative cartilage after autologous chondrocyte implantation: Preliminary results. Radiology, 239, 201-208. Watrin, A., Ruaud, J.-P. B., Olivier, P.T.A., Guigamp, N.C., Gonord, P.D., Netter, P.A., Blum, A.G., Guillot, G.M., Gillet, P.M., Loeuille, D.H., 2001. T2 mapping of rat cartilage. Radiology 219, 395-402. Welsch, G.H., Trattnig, S., Scheffler, K., Quirbach, S., Marlovits, S., Domayer, S., Bieri, O., Mamisch, T.C., 2008a. Magnetization transfer contrast and T2 mapping in the evaluation of cartilage repair tissue with 3T MRI. Journal of Magnetic Resonance Imaging 28, 979986. Welsch, G.H., Mamisch, T.C., Hughes, T., Zilkens, C., Quirbach, S., Scheffer, K., Kraff, O., Schweitzer, M.E., Szomolanyi, P., Trattnig, S., 2008b. In vivo biochemical 7.0 Tesla magnetic resonance: Preliminary results of dGEMRIC, zonal T2, and T2* mapping of articular cartilage. Investigative Radiology 43, 619-626 Welsch, G.H., Apprich, S., Zbyn, S., Mamisch, T.C., Mlynarik, V., Scheffler, K., Bieri, O., Trattnig S., 2011. Biochemical T2, T2* and magnetization transfer ratio MRI of the knee cartilage : feasibility at ultra-high field (7T) compared with high field (3T) strength. European Radiology 21, 1136-1143. Wheaton, A.J., Borthakur, A., Dodge, G.R., Kneeland, J.B., Schumacher, H.R., Reddy, R., 2004. Sodium magnetic resonance imaging of proteoglycan depletion in an in vivo model of osteoarthritis. Academic Radiology 11, 21-28. White, L.M., Sussman, M.S., Hurtig, M., Probyn, L., Tomlinson, G., Kandel, R., 2006. Cartilage T2 assessment: differentiation of normal hyaline cartilage and reparative tissue after arthroscopic cartilage repair in equine subjects. Radiology 241, 407-414. Wiewiorski, M., Miska, M., Kretzschmar, M., Studler, U., Bieri, O., Valderrabano, V., 2013. Delayed gadolinium-enhanced MRI of cartilage of the ankle joint: Results after autologous matrix-induced chondrogenesis (AMIC)-aided reconstruction of osteochondral lesions of the talus. Clinical Radiology 68, 1031-1038. Williams, A., Qian, Y., Bear, D., Chu, C.R., 2010. Assessing degeneration of human articular cartilage with ultra-short echo time (UTE) T2* mapping. Osteoarthritis and Cartilage 18, 539–546. Wucherer, K.L., Ober, C.P., Conzemius, M.G., 2012. The use of delayed gadolinium enhanced magnetic resonance imaging of cartilage and T2 mapping to evaluate articular cartilage in the normal canine elbow. Veterinary Radiology and Ultrasound 53, 57-63. Xia, Y., Farquhar, T., Burton-Wuster, N., Vernier-Singer, M., Lust, G., Jelinski, L., 1995. Self-diffusion monitors degraded cartilage. Archives of Biochemistry and Biophysics 10, 323-328.

Page 27 of 31

848 849 850 851 852 853 854 855

Xia Y., 1998. Relaxation anisotropy in cartilage by NMR microscopy (µMRI) at 14-µm resolution. Magnetic Resonance in Medicine 39, 941-949.

Zhou, J., van Zijl, P.C.M., 2006. Chemical exchange saturation transfer imaging and spectroscopy. Progress in Nuclear Magnetic Resonance Spectroscopy 48, 109-136.

Page 28 of 31

856

Figure legends

857 858

Fig. 1. Biochemical composition of cartilage. Articular cartilage is composed of hyaline

859

cartilage and calcified cartilage (CC) overlying the subchondral bone (SB). Hyaline cartilage

860

is comprised of three zones. In the superficial zone (SZ), which is in contact with the synovial

861

fluid (SF), chondrocytes (grey ellipses) are flattened and collagen fibres are parallel to the

862

articular surface. The transitional zone (TZ) is characterised by rounded chondrocytes and

863

arcade-like collagen fibres. In the radial zone (RZ), rounded chondrocytes are aligned in

864

columns and collagen fibres are perpendicular to the articular surface. Collagen fibres are

865

formed by aggregates of tropocollagen subunits. These are composed of a triple helix of

866

polypeptide chains. Although neutral overall, these chains bear a negative (plain arrow) and a

867

positive (dotted arrow) charge. A hydrated gel expands through the collagen network, made of

868

sulphated glycosaminoglycans (GAGs; black pins) linked to a core protein (brown line) to

869

form a proteoglycan (PG). Each PG is connected to HA (woven blue line) by a link protein

870

(orange circle).

871

GAGs are composed of hundreds of negatively charged units, counterbalanced by sodium.

872

Therefore, water exists in three forms: molecular water (linked to the negative [plain arrow]

873

or positive [dotted arrow] charges of collagen and GAG) and free water (attracted by the

874

sodium osmotic pressure; not shown in this figure). In cartilage, protons are found in the

875

hydrogen nucleus (H) of every molecule (collagen, water, GAGs).

876

Orange double arrows illustrate the chemical exchange saturation transfer (CEST) of protons

877

that is possible between amide and hydroxyl groups and is used in a specific compositional

878

imaging technique described in this review.

879

Page 29 of 31

880

Fig. 2. T1 rho (ρ). This sequence is characterised by the application of a 90° radiofrequency

881

pulse (B) followed by a spin-locking pulse (C). The magnetisation no longer relaxes

882

according to T2, but instead relaxes according to a time called T1ρ (D).

883 884

Fig. 3. Diffusion weighted imaging. In this technique a gradient pulse is applied, followed by

885

a 180° radiofrequency pulse and a second gradient pulse. Collagen fibres (green lines) restrict

886

the motion of water molecules (blue dots). In the cartilage radial zone, collagen fibres are

887

aligned perpendicular to articular surface. If the pair of gradient pulses is aligned with the

888

direction of the collagen fibres, the dephasing (blue arrow) of water by the first gradient pulse

889

is not compensated for by the effect (shorter backward blue arrow) of the 180°RF pulse and

890

the second gradient pulse, leading to a loss of signal. If the gradient pulse is perpendicular to

891

the direction of the collagen fibre, no loss of signal occurs, since the water molecule is in the

892

same position and is subject to the same pulse. The intensity of the gradient pulse is

893

represented by the greyscale arrow and the greyscale background under the collagen fibres.

894 895

Fig. 4. Glycosaminoglycan (GAG) chemical exchange saturation transfer (GAG-CEST;

896

adapted from Sherry and Woods, 2009 and Ling et al., 2009). A specific radiofrequency pulse

897

selectively activates GAG protons so that high energy protons and low energy protons are in

898

equal numbers. The system eventually becomes saturated; no energy transfer is possible and

899

there is no net magnetisation. The saturated protons of GAGs can exchange with the

900

unsaturated protons of water. This transfer induces a change in water magnetisation that can

901

be recorded.

902 903

Fig. 5. Example of a T2 mapping of the medial femoral and tibial condyles of the ovine knee.

904

The colour scale assigns a colour to each T2 relaxation time. Tissue with a long relaxation

Page 30 of 31

905

time (RT) will appear red-green. Short RT tissue will appear light blue, while very short RT

906

will appear dark blue.

907 908

Fig. 6. Example of a T1 mapping of the medial femoral and tibial condyles of the ovine knee

909

by delayed gadolinium-enhanced MR imaging of cartilage. T1 is measured in three different

910

regions of interest.

Page 31 of 31