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