- Email: [email protected]
Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study
Accepted Manuscript Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study Elisabeth Bandak, PT, MSc, Mikael Boesen, MD, PhD, Henning Bliddal, MD, DMSc, Robert GC. Riis, MD, Henrik Gudbergsen, MD, PhD, Marius Henriksen, PT, PhD PII:
S1063-4584(15)01190-5
DOI:
10.1016/j.joca.2015.05.032
Reference:
YJOCA 3508
To appear in:
Osteoarthritis and Cartilage
Received Date: 1 December 2014 Revised Date:
30 April 2015
Accepted Date: 26 May 2015
Please cite this article as: Bandak E, Boesen M, Bliddal H, Riis RG, Gudbergsen H, Henriksen M, Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study, Osteoarthritis and Cartilage (2015), doi: 10.1016/j.joca.2015.05.032. 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.
ACCEPTED MANUSCRIPT 1
Associations between muscle perfusion and symptoms in
2
knee osteoarthritis: a cross sectional study
RI PT
3 4
1
Elisabeth Bandak, PT, MSc, 1,2Mikael Boesen, MD, PhD, 1Henning Bliddal,MD, DMSc, 1,2Robert GC
5
Riis, MD, 1 Henrik Gudbergsen, MD, PhD, 1Marius Henriksen, PT, PhD
7
1
8
Frederiksberg, Copenhagen, Denmark
9
2
The Parker Institute, Dept. of Rheumatology, Copenhagen University Hospital, Bispebjerg &
M AN U
10
SC
6
Department of Radiology, Copenhagen University Hospital, Bispebjerg & Frederiksberg,
Copenhagen, Denmark
11
Email addresses
13
EB: [email protected]
14
MB: [email protected]
15
HB: [email protected]
16
RR: [email protected]
17
HG: [email protected]
18
MH: [email protected]
EP
AC C
19
TE D
12
20 21 22 1
ACCEPTED MANUSCRIPT Correspondence
24
Elisabeth Bandak
25
The Parker Institute
26
Bispebjerg & Frederiksberg Hospitals
27
Nordre Fasanvej 57
28
2000 Frederiksberg
29
Phone: +45 38164169
30
Fax: +45 38164159
31
E-mail: [email protected]
34
Abstract: 248
35
Manuscript text: 2732
36
Tables and figures: 5
37
References: 50
39 40 41
Running title: Muscle perfusion in knee osteoarthritis
AC C
38
M AN U
Word count
EP
33
TE D
32
SC
RI PT
23
42 43 44 2
ACCEPTED MANUSCRIPT Abstract
46
Objective: To investigate the association between muscle perfusion in the peri-articular knee
47
muscles assessed by dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) and
48
symptoms in patients with knee osteoarthritis (KOA).
49
Design: In a cross-sectional setting, muscle perfusion was quantified by DCE-MRI in KOA. Regions
50
of interest were drawn around the peri-articular muscles, summed and averaged into one single
51
“Total Muscle Volume” volume of interest (VOI). Symptoms were assessed via the Knee injury and
52
Osteoarthritis Outcome Score (KOOS) (0: worst; 100: best).
53
Results: DCE-MRI and clinical data were analyzed in 94 patients. The typical participant was a
54
woman with a mean age of 65 years, and a body mass index of 32 kg/m2. Reduced multiple
55
regression models analyzing the association between KOOS and DCE-MRI perfusion variables of
56
Total Muscle Volume showed a statistically significant association between Nvoxel% and KOOS pain
57
(0.41 (SE 0.14); p=0.0048). Nvoxel% was defined as the proportion of highly perfused voxels; i.e. the
58
voxels that show an early and rapid increase on the signal intensity versus time curves, reach a
59
plateau state (plateau pattern) and then showing a relatively rapid decline (washout pattern)
60
relative to the total number of voxels within the muscle VOI.
61
Conclusions: More widespread perfusion in the peri-articular knee muscles was associated with
62
less pain in patients with KOA. These results give rise to investigations of the effects of exercise on
63
muscle perfusion and its possible mediating role in the causal pathway between exercise and pain
64
improvements in the conservative management of KOA.
65
Keywords: Muscle perfusion; Magnetic resonance imaging, knee; osteoarthritis, symptoms
AC C
EP
TE D
M AN U
SC
RI PT
45
3
ACCEPTED MANUSCRIPT 66 67
Introduction Knee osteoarthritis (KOA) is considered a whole-joint disease, including affection of periarticular structures such as muscles and bursa1. Pain is the cardinal symptom in KOA2-4. The
69
pathology of pain has been a main target in understanding the disease with bone marrow lesions
70
(BMLs), synovitis4-6, and the infrapatellar fat pad (IPFP) suggested to play important roles in the
71
generation of pain in KOA7,8. KOA pain is predominantly localized to the joint lines, but patients
72
may also experience a more generalized and unspecific pattern of knee pain involving the whole
73
knee region including peri-articular structures such as muscles9-12.
SC
It has become evident that inflammation plays an important role in the pathogenesis of
M AN U
74
RI PT
68
osteoarthritis4,13,14, and both synovitis and BMLs have shown to be associated with pain and
76
structural progression in KOA15,16. The role of muscle tissue in the symptomatology of KOA is not
77
clear, but findings of inflammatory markers in peri-articular knee muscles in patients with KOA
78
suggest that inflammatory processes are not only confined to articular structures, and that
79
inflammatory processes in peri-articular muscles may contribute to the development of pain17,18.
80
In current practice, conventional radiographs (CR) remain the central component in
81
diagnosing KOA and assess structural progression19 although there is a poor association between
82
radiographic findings and clinical features19,20. This may be due to the fact that CR cannot capture
83
key elements of KOA pathology including inflammation and soft tissue pathology19-21. In contrast
84
to CRs, MRI provides a unique visualization of all the anatomical structures involved in KOA
85
including bone marrow and musculature21,22. The use of dynamic MRI sequences recorded prior to
86
and during intravenous (IV) bolus infusion of a Gadolinium (Gd) contrast medium (Dynamic
87
Contrast Enhanced MRI (DCE-MRI)23 can be used to extract perfusion parameters in both hand and
88
knee joints where the speed, degree, and distribution of the contrast uptake using time-intensity
AC C
EP
TE D
75
4
ACCEPTED MANUSCRIPT curves reflects the histological degree of inflammation in the tissue24 and the change in
90
enhancement over time can be used to assess treatment response in RA2526. DCE-MRI is a valid
91
method to describe musculoskeletal tumors, where high perfusion parameters correlate to higher
92
degree of histologically proven high-grade malignant tumor tissue 27. In KOA, DCE-MRI has only
93
been used in a few studies8,28, but associations between synovial perfusion and synovitis grading
94
assessed both macro- and microscopically have been described28, and recently higher perfusion of
95
the IPFP was associated with pain severity and function88.
SC
96
RI PT
89
When analyzing DCE-MRI images in KOA patients, we have observed that tissue enhancement (i.e. tissues with contrast agent uptake) is not confined to intra-articular structures
98
but also includes different patterns of contrast enhancement in the peri-articular knee muscle
99
tissue as well. It is unknown if the degree of enhancement in these muscles is pathological and
M AN U
97
contributes to the symptoms in KOA or if the contrast uptake seen is due to the basis perfusion of
101
skeletal muscle tissue29-31.
102
TE D
100
The purpose of this study was to investigate the association between muscle perfusion in the peri-articular knee muscles assessed by DCE-MRI and symptoms in patients with
104
KOA. We hypothesized that higher contrast enhancement, and thus perfusion in the peri-articular
105
muscles, would be associated with worse symptoms in patients with KOA. Understanding the role
106
of perfusion in the peri-articular knee muscles in KOA symptomatology may contribute to our
107
understanding of how exercise exerts its impact on symptomatic relief in KOA, which can be used
108
to optimize recommended exercise therapies.
AC C
EP
103
109
5
ACCEPTED MANUSCRIPT 110
Method
111
Study population Cross sectional data from a weight-loss maintenance study in obese patients with
113
radiographically verified tibiofemoral KOA were included in this study (the LIGHT study;
114
ClinicalTrials.gov: NCT00938808). Data used in the present study were gathered at the 1 year
115
follow-up. The LIGHT study was approved by the local health research ethical committee and
116
conducted in accordance with the Helsinki Declaration, as revised in 2000. Written and oral
117
informed consent was obtained from each patient. All participants had previously participated in a
118
68 week weight-loss study; the CAROT study (ClinicalTrials.gov identifier: NCT00655941). The main
119
inclusion criteria were age > 50 years, a body mass index (BMI) ≥ 30 kg/m2, and clinical KOA
120
diagnosis verified by radiographs obtained at baseline of the LIGHT study32. At inclusion, the
121
participants chose a target knee, defined as the most symptomatic knee, for all subsequent
122
assessments.
SC
M AN U
TE D
124
Dynamic Contrast Enhanced MRI protocol (DCE-MRI)
EP
123
RI PT
112
MRI of the target knee, i.e. the most symptomatic knee, was performed on a 3T Siemens
126
Verio® system. The patients were scanned in supine position using a dedicated 16-channel
127
send/receive knee coil. The MRI protocol used has been described in detail earlier8. As
128
recommended by the Danish Health and Medicines Authority, decreased renal function with an
129
estimated glomerular filtration rate (eGFR <60 ml/min/1.73m2) contraindicated administration of
130
IV contrast medium, and thereby DCE-MRI. The DCE-MRI sequence was always performed a
131
minimum of 25 minutes into the scan assuming a relaxed state of the patient with normalization
132
of potential perfusion changes induced by movement.
AC C
125
6
ACCEPTED MANUSCRIPT 133
Image analysis The DCE-MRI sequence was performed in the axial plane in all patients, covering the
135
suprapatellar recess to the insertion of the patellar ligament on the tibia. All DCE-MRI analyses
136
were performed by an investigator blinded to the clinical data (EB) supervised by MB and RR. DCE-
137
MRI slices were analyzed using the computer software Dynamika® version 2.4.4.1 (Image Analysis
138
LTD, London, http://www.imageanalysis.org.uk)33,34. Motion correction between temporal slices
139
was applied on all the available axial DCE-MRI slices35 before regions of interest (ROIs) were drawn
140
around each of the following muscles: vastus lateralis, vastus medialis, biceps femoris, sartorius,
141
gracilis, triceps surae, popliteus, semitendinosus and semimembranosus. Major vascular branches
142
were avoided when drawing the ROIs (Figure 1, A). All the drawn ROIs were summed and averaged
143
into one single “Total Muscle Volume” VOI. Osirix® v. 5.7.1 was used to confirm the anatomical
144
boundaries of the muscles.
SC
M AN U
As previously described in detail 34 ; signal intensity versus time curves can be generated
TE D
145
RI PT
134
for each voxel within a ROI by using appropriate software. Dynamika® uses a robust classification
147
scheme34 where each signal intensity versus time curves automatically are assigned to one of the
148
following enhancement patterns: Tissues with high perfusion are likely to show an early and rapid
149
increase on the signal intensity versus time curves, reach a plateau state (plateau pattern) and
150
then potentially show a relatively rapid decline (washout pattern), whereas tissues with lower
151
perfusion showing a slower increase and potentially do not reach a plateau state (persistent
152
pattern), and tissue with no contrast uptake (no enhancement pattern)25,34. The voxels are
153
automatically assigned and color coded in a Gadolinium (Gd)-enhancement map to one of the four
154
enhancement patterns: washout (red), persistent (blue), plateau (green), or no enhancement (no
155
color)34 26(Figure 1, B).
AC C
EP
146
7
ACCEPTED MANUSCRIPT 156
Various DCE-MRI perfusion variables can be extracted from the signal intensity versus time curves, the most relevant in the setting of musculoskeletal imaging being: i) maximal
158
enhancement (ME), calculated as the maximum signal intensity relative to the baseline signal
159
intensity; ii) the Initial Rate of Enhancement (IRE) i.e. the speed of enhancement or upslope on the
160
signal intensity versus time curve calculated as the relative increase in signal intensity per
161
second(%/s) from time of enhancement onset (in seconds after the contrast injection) until ME is
162
reached, and iii) Nvoxel, the sum of voxels with plateau or washout enhancement patterns, i.e. the
163
highest perfused voxels 24,34,36.
SC
RI PT
157
Values and maps of IRE and ME were generated, color-coded and superimposed on the
165
grey scale dynamic images. The highest values for IRE and ME were shown in bright yellow and
166
lower values in a spectrum of red colors (Figure 1, C-D). Intra- and inter-observer reliability
167
analysis for the Total Muscle Volume (cm3) was performed on a random, repeated subsample
168
(n=10) with a minimum of 4 weeks between drawings using single measures intraclass correlation
169
coefficients (ICC) as a measure of relative reliability and measurement errors (ME) calculated as
170
the square root of the residual mean square values obtained from analyses of variance, as a
171
measure of absolute reliability. The intra-observer reliability of the total muscle volume (cm3) was
172
ICC=0.99 and ME= 6.5 cm3; inter-observer reliability: ICC=0.96 and ME=10.8 cm3. The intra-
173
observer reliability of the IRExNvoxel was ICC=0.97 and ME= 23.3; inter-observer reliability: ICC=0.71
174
and ME=65.0. The intra-observer reliability of the MExNvoxel was ICC=0.99 and ME= 2613.6; inter-
175
observer reliability: ICC=0.91 and ME=7312.1. The intra-observer reliability of the Nvoxel% was
176
ICC=0.99 and ME=1.0%; inter-observer reliability: ICC=0.94 and ME=5.6%.
AC C
EP
TE D
M AN U
164
8
ACCEPTED MANUSCRIPT 177 178 179
Outcome measures The radiographs were scored by a trained radiologist (MB) according to the KellgrenLawrence (KL) grading of radiographic disease severity37. To assess the patients’ symptoms we applied the Knee injury and Osteoarthritis Outcome
181
Score (KOOS)38 questionnaire. KOOS is a self-administered patient-reported outcome measure,
182
assessing five domains of importance to patients with KOA (Pain, Symptoms, Function in daily
183
living, Function in sport and recreation, and Knee related quality of life). Each domain includes
184
individual items answered on 0-4 Likert scale. A normalized 0-100 score (0 indicating the worst,
185
100 indicating the best) is calculated for all domains38. The items in the KOOS questionnaire relate
186
to target knee symptoms during the previous week. In this study we used a Danish touch screen
187
version of KOOS39.
SC
M AN U
188
RI PT
180
The DCE-MRI perfusion variables of the Total Muscle Volume VOI were as follows: IRExNvoxel, MExNvoxel, and Nvoxel%, where IRE and ME were quantified as average measures of the
190
whole segmented region. By multiplying Nvoxel (the sum of voxels with plateau or washout
191
enhancement patterns, i.e. the highest perfused voxels) with the means of the IRE and ME the
192
variables become composite parameters reflecting both the volume and the degree of higher
193
perfused voxels (IRExNvoxel, MExNvoxel)24. Nvoxel% is the proportion of highly perfused voxels
194
reaching either a plateau or washout enhancement patterns relative to the total number of voxels
195
within the muscle VOI. Nvoxel, IRExNvoxel, and MExNvoxel are all variables obtained from DCE-MRI that
196
have been used in previous studies8,24,26,36.
AC C
EP
TE D
189
197
9
ACCEPTED MANUSCRIPT 198
Statistical methods The analyses were pre-specified and conducted on all participants with valid DCE-MRI data,
200
i.e. no imputation for missing data (Figure 2, Flowchart). Five multiple regression models were
201
built to determine the association between the 5 dependent variables (the 5 KOOS subscales and
202
KL score), each with the 3 DCE-MRI perfusion variables of the Total Muscle Volume: IRExNvoxel,
203
MExNvoxel, and Nvoxel% as independent variables. Reductions of the models were done manually
204
eliminating
205
coefficients/slopes) one at a time based on p-values (highest eliminated first). Reductions were
206
only done in regression models that were overall statistically significant (p<0.05). The reduced
207
models were repeated adjusting for possible confounders pre-defined as sex, age, and BMI. To
208
assess the adequacy of the regression model(s) describing the observed data we investigated the
209
model features via the predicted values and the studentized residuals; i.e. the residuals had to be
210
normally distributed (around zero), and be independent of the predicted values. The analyses
211
were done using SAS software, version 9.3 (SAS Inc., Cary, NC, USA). Statistical significance was
212
accepted at p<0.05.
non-significant
(p>0.05)
variables
(i.e.
non-significant
beta-
EP
213
TE D
M AN U
SC
statistically
RI PT
199
Results
215
Patient characteristics
AC C
214
216
In this cross-sectional study, all 134 patients completed the 1-year visit in the LIGHT study.
217
Of these, MRI was performed with 97 patients. Reasons for missing MRI assessment were due to
218
either: (1) the target knee was above 53 cm in circumference and could therefore not fit into the
219
send/receive knee coil (2) decreased renal function with an eGFR < 60 ml/min/1.73m2. Technical
220
difficulties with the DCE-MRI analysis occurred in 3 participants. Thereby the number of analyzed 10
ACCEPTED MANUSCRIPT DCE-MRI data sets was 94 (70% of the study population) (Figure 2). Patient characteristics are
222
available in Table 1. The typical patient participating in this study was a woman aged 65 years
223
with a BMI of 32 kg/m2. The 94 patients who were analyzed in this study had a lower weight and
224
BMI compared with the 40 patients without completed or legible MRIs (Supplementary file).
225
Outcomes
RI PT
221
Summaries of the regression analyses are provided in Table 2. The reduced multiple
227
regression models analyzing the association between KOOS pain and the DCE-MRI perfusion
228
variables of the Total Muscle Volume showed a statistically significant association between Nvoxel%
229
and KOOS pain (Table 2). Adjustment for sex, age, and BMI did not change the results significantly.
230
There were no statistically significant associations between the other KOOS subdomains
231
(Symptoms, Function in daily living, Function in sport and recreation, and Knee related quality of
232
life),the KL score and the DCE-MRI muscle perfusion variables.
233 234
Discussion
TE D
M AN U
SC
226
In this cross sectional study we found a significant association between pain and the
236
proportion of the peri-articular knee muscle tissue with a distinct perfusion pattern among
237
patients with KOA assessed by DCE-MRI. The association suggests that for each percentage point
238
increment in Nvoxel%, the KOOS pain score is 0.41 higher (less pain) in patients with KOA. Thus,
239
individuals with a larger proportion of the peri-articular muscles showing a plateau or washout
240
contrast uptake pattern seemed to have less pain. By consequence, individuals with larger areas
241
with no enhancement or persistent contrast uptake patterns had more pain. IRExNvoxel and
242
MExNvoxel were not associated with pain. We found no associations between the other KOOS
243
subdomains (Symptoms, Function in daily living, Function in sport and recreation, and Knee
AC C
EP
235
11
ACCEPTED MANUSCRIPT 244
related quality of life), the KL score, and the DCE-MRI muscle perfusion variables in the peri-
245
articular knee muscles. These results are in contrast to our hypothesis and are surprising. To the best of our
247
knowledge this is the first study to assess muscle perfusion in patients with KOA, so direct
248
comparisons with other studies are not possible. However, our results are indirectly supported by
249
the fact that muscle weakness is a typical feature of KOA40, and that exercise has beneficial effects
250
on pain in KOA41. Further, exercise has been shown to increase muscle perfusion in both young
251
and elderly healthy subjects42,43, and it is therefore possible that the variation in muscle perfusion
252
in our study reflects variations in the physical fitness status of the cohort. In further (indirect)
253
support of this notion, previous results of the effects of weight loss in this cohort showed that
254
increased muscle strength was associated with decreased pain44.
M AN U
SC
RI PT
246
Our findings are of potential importance in explaining the underlying analgesic mechanisms
256
of exercise interventions in KOA. Several studies have assessed possible theories, including neural
257
pain processing45, biomechanical40,46 and anti-inflammatory mechanisms8,47,48. As the muscles are
258
the tissue primarily affected by exercise, it is relevant to investigate if muscle perfusion is
259
associated with pain in KOA. Future longitudinal studies are wanted, and should investigate if
260
changes in muscle perfusion are a contributing factor explaining the beneficial outcomes of
261
exercise, and if specific types of exercise have specific effects on the muscle perfusion.
AC C
EP
TE D
255
262
Our results indicate that previous findings of a relationship between high pain and high
263
perfusion of the synovium and the IPFP in KOA patients may not be accompanied by similar
264
changes in peri-articular structures8,28. Inflammatory substances have been detected in the
265
quadriceps muscles of individuals with KOA17,28,49. However, these were not associated with pain.
266
Hyperperfusion in articular structures is generally recognized as a sign of inflammation24-26, and as 12
ACCEPTED MANUSCRIPT 267
such undesirable and a potential treatment target in KOA. In contrast, the present results indicate
268
that higher muscle perfusion may be beneficial in KOA and could possibly be a parallel to the
269
changes in muscle capillarization following exercise as observed in other studies43,50. This study has some limitations associated with our cross-sectional design, which
271
precludes an evaluation of a possible relationship between changes in symptoms and DCE-MRI
272
perfusion variables. Also, as the underlying MRI protocol focused on the knee joint, the images
273
were limited to the area of the send/receive knee coil, therefore we were only able to analyze
274
certain parts of the muscles. Further, the likelihood of a false positive finding is not negligible due
275
to the multiple statistical tests. However, the results seem robust even after adjustment for
276
covariates. With a Bonferroni adjustment for multiple tests (6 regressions) the threshold of
277
statistical significance is 0.0083. The statistically significant regression results (KOOS pain; Table 2)
278
respect this threshold for significance, except the covariate adjusted model. Important strengths
279
to our study include the relative large study sample, the use of a 3T MRI system, and the pre-
280
specified analysis plan.
282
Conclusion
EP
281
TE D
M AN U
SC
RI PT
270
In conclusion, this study showed, in a cross-sectional setting, that more widespread
284
perfused voxels in the peri-articular knee muscles was associated with less pain in patients with
285
KOA. These results give rise to investigations of the effects of exercise on muscle perfusion and its
286
possible mediating role in the causal pathway between exercise and pain improvements in the
287
conservative management of KOA.
AC C
283
13
ACCEPTED MANUSCRIPT 288
290 291
Acknowledgements We would like to thank the staff at the Department of Radiology, Bispebjerg and Frederiksberg Hospitals for helping in obtaining the MRI.
RI PT
289
292 293
Financial support
This study was supported by grants from The Danish Physiotherapy Association, The
295
Capital Region of Copenhagen, Bispebjerg-and Frederiksberg Hospitals, The Danish Rheumatism
296
Association, The Oak Foundation, The Velux Foundation, The Cambridge Weight plan UK, The
297
Augustinus Foundation, The A.P.Møller Foundation for the Advancement of Medical Science,
298
Hørslev Fonden, Bjarne Jensens Fond and Aase og Ejnar Danielsens fond.
300
M AN U
The funding sources did not influence study design, collection, analysis and interpretation of data, writing, or the decision to submit the manuscript for publication.
301
304 305 306 307 308
EP
303
Author contributions
EB contributed to the study design, statistical design and analysis, interpretation of the data, drafted and revised the manuscript, and approved the final version.
AC C
302
TE D
299
SC
294
MB contributed to the study design, radiological supervision, and interpretation of the data, revised the manuscript, and approved the final version. HB contributed to the study design, interpretation of the data, revised the manuscript, and approved the final version.
14
ACCEPTED MANUSCRIPT
311 312 313 314 315 316
interpretation of the data, revised the manuscript, and approved the final version. HG contributed to interpretation of the data, revised the manuscript, and approved the final version. MH contributed to the study design, statistical design and analysis, interpretation of the data, revised the manuscript, and approved the final version.
All the authors had access to the data and analyses. EB takes responsibility for the integrity of the work as a whole, from inception to finished article.
M AN U
317 318
RI PT
310
RR contributed in the inter-observer reliability analysis, radiological supervision, and
SC
309
Conflicts of interest
MB reports grants from The Oak Foundation, grants from Abbvie A/S Denmark, non-
320
financial support from ESAOTE MRI vendor, Genoa Italy, non-financial support from Image Analysis
321
LTD London UK, outside the submitted work; All other authors declare no conflicts of interest.
EP AC C
322
TE D
319
15
ACCEPTED MANUSCRIPT
324 325
326 327
References (1) Loeser R, Goldring S, Scanzello C, Goldring MB. Osteoarthritis: a disease of the joint as an organ. Arthritis & Rheumatism 2012;64(6):1697-707.
(2) Schaible HG. Mechanisms of chronic pain in osteoarthritis. Curr Rheumatol Rep 2012
RI PT
323
Dec;14(6):549-56.
(3) Felson DT. The sources of pain in knee osteoarthritis. Curr Opin Rheumatol 2005 Sep;17(5):624-8.
329
(4) Bijlsma JWJ. Osteoarthritis: an update with relevance for clinical practice. Lancet
331
2011;377(9783):2115-26.
M AN U
330
SC
328
(5) Zhang Y, Nevitt M, Niu J, Lewis C, Torner J, Guermazi A, et al. Fluctuation of knee pain and changes
332
in bone marrow lesions, effusions, and synovitis on magnetic resonance imaging. Arthritis Rheum
333
2011 Mar;63(3):691-9.
(6) Guermazi A, Roemer FW, Hayashi D, Crema MD, Niu J, Zhang Y, et al. Assessment of synovitis with
TE D
334
contrast-enhanced MRI using a whole-joint semiquantitative scoring system in people with, or at
336
high risk of, knee osteoarthritis: the MOST study. Ann Rheum Dis 2011 May;70(5):805-11.
337
(7) Crema MD, Felson DT, Roemer FW, Niu J, Marra MD, Zhang Y, et al. Peripatellar synovitis:
EP
335
comparison between non-contrast-enhanced and contrast-enhanced MRI and association with
339
pain. The MOST study. Osteoarthritis Cartilage 2013 Mar;21(3):413-8.
340
AC C
338
(8) Ballegaard C, Riis RG, Bliddal H, Christensen R, Henriksen M, Bartels EM, et al. Knee pain and
341
inflammation in the infrapatellar fat pad estimated by conventional and dynamic contrast-
342
enhanced magnetic resonance imaging in obese patients with osteoarthritis: A cross-sectional
343
study. Osteoarthritis and Cartilage(0). 16
ACCEPTED MANUSCRIPT 344
(9) Thompson LR, Boudreau R, Hannon MJ, Newman AB, Chu CR, Jansen M, et al. The knee pain map:
345
Reliability of a method to identify knee pain location and pattern. Arthritis & Rheumatism 2009 Jun
346
15;61(6):725-31.
350
351
RI PT
349
osteoarthritis of the knee. Osteoarthritis and Cartilage 1998 Sep;6(5):318-23.
(11) McLean SM, Burton M, Bradley L, Littlewood C. Interventions for enhancing adherence with physiotherapy: A systematic review. Manual Therapy 2010 Dec;15(6):514-21.
SC
348
(10) Creamer P, Lethbridge-Cejku M, Hochberg MC. Where does it hurt? Pain localization in
(12) Wood LRJ, Peat G, Thomas E, Duncan R. Knee osteoarthritis in community-dwelling older adults:
M AN U
347
352
are there characteristic patterns of pain location? Osteoarthritis and Cartilage 2007 Jun;15(6):615-
353
23.
(13) Goldring M. Inflammation in osteoarthritis. Current opinion in rheumatology 2011;23(5):471-8.
355
(14) Berenbaum F. Osteoarthritis as an inflammatory disease (osteoarthritis is not osteoarthrosis!).
357
Osteoarthritis and Cartilage 2013 Jan;21(1):16-21.
(15) Hill C, Hunter D, Clancy M, Guermazi A, Genant H, Gale D, et al. Synovitis detected on magnetic
EP
356
TE D
354
resonance imaging and its relation to pain and cartilage loss in knee osteoarthritis. Annals of the
359
Rheumatic Diseases 2007;66(12):1599-603.
360
AC C
358
(16) Tanamas SK, Wluka AE, Pelletier JP, Pelletier JM, Abram F, Berry PA, et al. Bone marrow lesions in
361
people with knee osteoarthritis predict progression of disease and joint replacement: a longitudinal
362
study. Rheumatology (Oxford) 2010 Dec;49(12):2413-9.
17
ACCEPTED MANUSCRIPT 363
(17) Levinger I, Levinger P, Trenerry M, Feller JA, Bartlett JR, Bergman N, et al. Increased inflammatory
364
cytokine expression in the vastus lateralis of patients with knee osteoarthritis. Arthritis &
365
Rheumatism 2011;63(5):1343-8.
(18) Oliveria SA, Felson DT, Cirillo PA, Reed JI, Walker AM. Body weight, body mass index, and incident
367
symptomatic osteoarthritis of the hand, hip, and knee. Epidemiology 1999 Mar;10(2):161-6.
368
(19) Bliddal H, Boesen M, Christensen R, Kubassova O, Torp-Pedersen S. Imaging as a follow-up tool in
371
SC
370
clinical trials and clinical practice. Best Pract Res Clin Rheumatol 2008 Dec;22(6):1109-26.
(20) Bedson J, Croft PR. The discordance between clinical and radiographic knee osteoarthritis: a
M AN U
369
RI PT
366
systematic search and summary of the literature. BMC Musculoskelet Disord 2008;9:116.
(21) Guermazi A, Roemer FW, Burstein D, Hayashi D. Why radiography should no longer be considered a
373
surrogate outcome measure for longitudinal assessment of cartilage in knee osteoarthritis. Arthritis
374
Res Ther 2011;13(6):247.
TE D
372
(22) Menashe L, Hirko K, Losina E, Kloppenburg M, Zhang W, Li L, et al. The diagnostic performance of
376
MRI in osteoarthritis: a systematic review and meta-analysis. Osteoarthritis Cartilage 2012
377
Jan;20(1):13-21.
(23) Verstraete KL, Van der Woude HJ, Hogendoorn PC, De-Deene Y, Kunnen M, Bloem JL. Dynamic
AC C
378
EP
375
379
contrast-enhanced MR imaging of musculoskeletal tumors: basic principles and clinical applications.
380
J Magn Reson Imaging 1996 Mar;6(2):311-21.
381
(24) Axelsen MB, Poggenborg RP, Stoltenberg M, Kubassova O, Boesen M, Horslev-Petersen K, et al.
382
Reliability and responsiveness of dynamic contrast-enhanced magnetic resonance imaging in
383
rheumatoid arthritis. Scand J Rheumatol 2013;42(2):115-22.
18
ACCEPTED MANUSCRIPT 384
(25) Kubassova O, Boesen M, Peloschek P, Langs G, Cimmino MA, Bliddal H, et al. Quantifying Disease
385
Activity and Damage by Imaging in Rheumatoid Arthritis and Osteoarthritis. Annals of the New York
386
Academy of Sciences 2009 Feb 15;1154:207-38.
(26) Boesen M, Kubassova O, Cimmino MA, Ostergaard M, Taylor P, Danneskiold-Samsoe B, et al.
RI PT
387
Dynamic Contrast Enhanced MRI Can Monitor the Very Early Inflammatory Treatment Response
389
upon Intra-Articular Steroid Injection in the Knee Joint: A Case Report with Review of the Literature.
390
Arthritis 2011;2011:578252.
392
393
(27) Nygren AT, Greitz D, Kaijser L. Skeletal muscle perfusion during exercise using Gd-DTPA bolus detection. J Cardiovasc Magn Reson 2000;2(4):263-70.
M AN U
391
SC
388
(28) Loeuille D, Rat AC, Goebel JC, Champigneulle J, Blum A, Netter P, et al. Magnetic resonance imaging in osteoarthritis: which method best reflects synovial membrane inflammation?: Correlations with
395
clinical, macroscopic and microscopic features. Osteoarthritis and Cartilage 2009 Sep;17(9):1186-
396
92.
397
TE D
394
(29) Heinonen I, Saltin B, Kemppainen J, Sipila HT, Oikonen V, Nuutila P, et al. Skeletal muscle blood flow and oxygen uptake at rest and during exercise in humans: a pet study with nitric oxide and
399
cyclooxygenase inhibition. Am J Physiol Heart Circ Physiol 2011 Apr;300(4):H1510-H1517.
400
(30) Levy M. Special Circulations. In: Berne R, Levy M, Koeppen B, Stanton B, editors. Physiology. Fifth
402
AC C
401
EP
398
Edition ed. Missouri: Mosby; 2004. p. 413-32.
(31) Bajwa A, Wesolowski R, Patel A, Saha P, Ludwinski F, Smith A, et al. Assessment of tissue perfusion
403
in the lower limb: current methods and techniques under development. Circ Cardiovasc Imaging
404
2014 Sep;7(5):836-43.
19
ACCEPTED MANUSCRIPT 405
(32) Riecke BF, Christensen R, Christensen P, Leeds AR, Boesen M, Lohmander LS, et al. Comparing two
406
low-energy diets for the treatment of knee osteoarthritis symptoms in obese patients: a pragmatic
407
randomized clinical trial. Osteoarthritis and Cartilage 2010 Jun;18(6):746-54.
(33) Kubassova O, Boesen M, Boyle RD, Cimmino MA, Jensen KE, Bliddal H, et al. Fast and robust
RI PT
408 409
analysis of dynamic contrast enhanced MRI datasets. Med Image Comput Comput Assist Interv
410
2007;10(Pt 2):261-9.
(34) Kubassova O, Boesen M, Cimmino MA, Bliddal H. A computer-aided detection system for
SC
411
rheumatoid arthritis MRI data interpretation and quantification of synovial activity. European
413
Journal of Radiology 2010;74(3):67-72.
414
M AN U
412
(35) Boesen M, Kubassova O, Parodi M, Bliddal H, Innocenti S, Garlaschi G, et al. Comparison of the manual and computer-aided techniques for evaluation of wrist synovitis using dynamic contrast-
416
enhanced MRI on a dedicated scanner. European Journal of Radiology 2011 Feb;77(2):202-6.
417
(36) Boesen M, Kubassova O, Bouert R, Axelsen M, Ostergaard M, Cimmino MA, et al. Correlation
TE D
415
between computer-aided dynamic gadolinium-enhanced MRI assessment of inflammation and
419
semi-quantitative synovitis and bone marrow oedema scores of the wrist in patients with
420
rheumatoid arthritis-a cohort study. Rheumatology 2012;51(1):134-43.
422
423
AC C
421
EP
418
(37) Kelgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Annals of the Rheumatic Diseases 1957;16(4):494-502.
(38) Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD. Knee Injury and Osteoarthritis Outcome
424
Score (KOOS)--development of a self-administered outcome measure. J Orthop Sports Phys Ther
425
1998 Aug;28(2):88-96.
20
ACCEPTED MANUSCRIPT 426
(39) Gudbergsen H, Bartels E, Krusager P, Waehrens E, Christensen R, nneskiold-Samsoe B, et al. Testretest of computerized health status questionnaires frequently used in the monitoring of knee
428
osteoarthritis: a randomized crossover trial. BMC Musculoskeletal Disorders 2011;12(1):190.
429
(40) Bennell KL, Wrigley TV, Hunt MA, Lim BW, Hinman RS. Update on the Role of Muscle in the Genesis
RI PT
427
430
and Management of Knee Osteoarthritis. Rheumatic Disease Clinics of North America 2013
431
Feb;39(1):145-76.
(41) McAlindon TE, Bannuru RR, Sullivan MC, Arden NK, Berenbaum F, Bierma-Zeinstra SM, et al. OARSI
SC
432
guidelines for the non-surgical management of knee osteoarthritis. Osteoarthritis and Cartilage
434
2014 Mar;22(3):363-88.
436
437 438
439
(42) Harris BA. The influence of endurance and resistance exercise on muscle capillarization in the elderly: a review. Acta Physiol Scand 2005 Oct;185(2):89-97.
(43) Hoier B, Hellsten Y. Exercise-induced capillary growth in human skeletal muscle and the dynamics
TE D
435
M AN U
433
of VEGF. Microcirculation 2014 May;21(4):301-14.
(44) Henriksen M, Christensen R, Danneskiold-Samsoe B, Bliddal H. Changes in lower extremity muscle mass and muscle strength after weight loss in obese patients with knee osteoarthritis: a
441
prospective cohort study. Arthritis Rheum 2012 Feb;64(2):438-42.
AC C
442
EP
440
(45) Henriksen M, Klokker L, Graven-Nielsen T, Bartholdy C, Jorgensen TS, Bandak E, et al. Exercise
443
therapy reduces pain sensitivity in patients with knee osteoarthritis: A randomized controlled trial.
444
Arthritis Care Res (Hoboken ) 2014 Jun 6.
21
ACCEPTED MANUSCRIPT 445
(46) Bennell KL, Hunt MA, Wrigley TV, Hunter DJ, McManus FJ, Hodges PW, et al. Hip strengthening
446
reduces symptoms but not knee load in people with medial knee osteoarthritis and varus
447
malalignment: a randomised controlled trial. Osteoarthritis Cartilage 2010 May;18(5):621-8.
(47) Helmark I, Mikkelsen U, Børglum J, Rothe A, Petersen M, Andersen O, et al. Exercise increases
RI PT
448 449
interleukin-10 levels both intraarticularly and peri-synovially in patients with knee osteoarthritis: a
450
randomized controlled trial. Arthritis research & therapy 2010;12(4):R126.
SC
452
(48) Petersen AM, Pedersen BK. The anti-inflammatory effect of exercise. Journal of applied physiology 2005 Apr 1;98(4):1154-62.
M AN U
451
453
(49) Levinger P, Caldow MK, Feller JA, Bartlett JR, Bergman NR, McKenna MJ, et al. Association between
454
skeletal muscle inflammatory markers and walking pattern in people with knee osteoarthritis.
455
Arthritis & Rheumatism 2011;63(12):1715-21.
(50) Weber MA, Hildebrandt W, Schröder L, Kinscherf R, Krix M, Bachert P, et al. Concentric resistance
457
training increases muscle strength without affecting microcirculation. European Journal of
458
Radiology 2010 Mar;73(3):614-21.
461 462 463
EP
460
AC C
459
TE D
456
464
22
ACCEPTED MANUSCRIPT Table and figure legends
466
Figure 1
467
Maps of DCE-MRI perfusion variables in a representative patient with KOA.
468
Axial dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) gradient echo (GRE) T1
469
w VIBE sequence approximately at the level of basis patella in a representative patient with KOA
470
(A-D).
471
A, The drawing of region of interests (ROIs) around of all the peri-articular knee muscles avoiding
472
major vascular branches.
473
B, Contrast Uptake Pattern: The voxels are automatically assigned and color coded in a Gadolinium
474
(Gd)-enhancement map to one of the four enhancement patterns: washout (red), persistent
475
(blue), plateau (green), or no enhancement (no color). C, maps of Initial Rate of Enhancement
476
(IRE), and D, maps of Maximum Enhancement (ME). Values of IRE and ME are color-coded and
477
superimposed on the grey scale dynamic images. The highest values are shown in bright yellow to
478
white and lower values in a spectrum of red colors.
479
White arrows: The popliteal artery. Blue arrows: Enhancement in the synovium in the retro-
480
patellar space.
AC C
EP
TE D
M AN U
SC
RI PT
465
481
482
Figure 2
483
Flowchart
484 23
ACCEPTED MANUSCRIPT Table 1
486
Demographics, self-reported symptoms, and imaging variables from the 94 patients with KOA
487
included in this study.
RI PT
485
488
Abbreviations: BMI, Body mass index; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL,
490
Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and
491
recreation; KL, Kellgren-Lawrence, DCE-MRI, dynamic contrast enhanced magnetic resonance
492
imaging;
493
Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of
494
the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and
495
semimembranosus muscles; ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel,
496
the sum of voxels with a plateau or washout enhancement pattern. Nvoxel%, the proportion of
497
voxels reaching either a plateau or washout enhancement pattern relative to the total number of
498
voxels within the muscle VOI.
499
*: Data from 93 patients.
500 501 502
AC C
EP
TE D
M AN U
SC
489
503
Table 2
504
Multiple regression analysis with KOOS subscales and KL score as dependent variables and DCE-
505
MRI perfusion variables of Total Muscle Volume as independent variables, (n=94).
506 24
ACCEPTED MANUSCRIPT Estimate: slope (SE); p-value.
508
Abbreviations: DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle
509
Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus
510
lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and
511
semimembranosus muscles.
512
ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a
513
plateau or washout enhancement pattern; Nvoxel%, the proportion of voxels reaching either a
514
plateau or washout enhancement pattern relative to the total voxel number of voxels within the
515
muscle VOI; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living;
516
QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation, KL, Kellgren-
517
Lawrence.
518
*Data from 93 patients.
TE D
519
M AN U
SC
RI PT
507
Supplementary Table
521
Demographics, clinical, and imaging variables of the group that completed the MRI and the group
522
without a completed or legible MRI.
AC C
523
EP
520
524
Abbreviations: MRI, Magnetic resonance imaging; BMI, Body mass index; KOOS, Knee injury and
525
Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related Quality of life;
526
Sport/Rec, Function in sport and recreation; KL, Kellgren-Lawrence.
527
¤: Data from 39 patients. *: Data from 93 patients.
528 25
ACCEPTED MANUSCRIPT Table 1. Demographics, self-reported symptoms, and imaging variables from the 94 patients with KOA included in this study.
77 (82 %) 65.3 ± 6.5 (52.4-77.4) 165.9 ± 8.1 (148.0-189.0) 88.9 ± 12.6 (68.3-130.2) 32.3 ± 3.7 (25.2-44.2)
M AN U
SC
62.9 ± 17.4 (22.2-100.0) 64.3 ± 19.0 (10.7-100.0) 67.6 ± 17.2 (17.6-98.5) 45.0 ± 18.9 (6.3-93.8) 30.7 ± 23.1 (0.0-95.0) 2.5 ± 0.8 (1-4) 0 (0 %) 10 (10.75 %) 36 (38.7 %) 38 (40.86 %) 9 (9.67 %)
EP
TE D
Demographics and clinical characteristics Female, no (%) Age (years) Height (cm) Weight (kg) BMI (kg/m2) KOOS scores Pain Symptoms ADL QOL Sport/Rec X-ray KL score* KL score, no. (0-4) 0, no. (%) 1, no. (%) 2, no. (%) 3, no. (%) 4, no. (%) DCE-MRI perfusion variables Total Muscle Volume Mean IRExNvoxel Mean MExNvoxel Nvoxel%
RI PT
Mean ± SD (min-max)
218.8 ± 119.7 (50.3-663.1) 74,172.3 ± 21,061.4 (23,233.7-12,8382.4) 74.9 ± 12.2 (17.6-95.3)
AC C
Abbreviations: BMI, Body mass index; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation; KL, KellgrenLawrence, DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and semimembranosus muscles; ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a plateau or washout enhancement pattern. Nvoxel%, the proportion of voxels reaching either a plateau or washout enhancement pattern relative to the total number of voxels within the muscle VOI. *: Data from 93 patients.
ACCEPTED MANUSCRIPT Table 2 Multiple regression analysis with KOOS subscales and KL score as dependent variables and DCE-MRI perfusion variables of Total Muscle Volume as independent variables, (n=94). Dependent variable
Independent variables
Model
Reduced model 1
KOOS pain
IRExNvoxel
-0.016 (0.019); p=0.4081
-0.011 (0.016); p=0.4691
-
-
MExNvoxel
0.000052 (0.00013); p=0.6805
-
-
-
Nvoxel%
0.41 (0.18), p=0.0258
IRExNvoxel
-0.016 (0.021); p=0.4380
MExNvoxel
0.00016 (0.00014); p=0.2584
Nvoxel%
0.22 (0.20); p=0.2766
IRExNvoxel
-0.020 (0.019); p=0.2908
MExNvoxel
0.00010 (0.00013); p=0.4172
Nvoxel%
0.26 (0.18); p=0.1505
IRExNvoxel
KL score*
RI PT -
-
0.00015 (0.00016); p= 0.3496
-
-
0.22 (0.24); p= 0.3568
-
-
-0.015 (0.02); p= 0.4491
-
-
0.00015 (0.00014); p= 0.2905
-
-
0.15 (0.21); p= 0.4851
-0.014 (0.021); p=0.4997
-
-
-0.008 (0.022); p=0.7168
MExNvoxel
0.00013 (0.00014); p=0.3443
-
-
0.00016 (0.00016); p=0.3170
Nvoxel%
0.11 (0.20); p=0.5826
-
-
0.03 (0.24) p= 0.8884
IRExNvoxel
-0.008 (0.026); p=0.7560
-
-
0.004 (0.027); p= 0.8918
MExNvoxel
0.00018 (0.00017); p=0.2960
-
-
0.00021 (0.00019); p= 0.2756
Nvoxel%
-0.09 (0.25); p=0.7132
-
-
-0.20 (0.29): p= 0.4870
IRExNvoxel
0.000 (0.001); p=0.5032
-
-
0.000 (0,001); p=0.773
MExNvoxel
0.00000 (0.00001); p=0.6399
-
-
0.00000 (0.00001); p=0.9082
Nvoxel%
0.00 (0.01); p=0.7829
-
-
0.00 (0.01); p= 0.6998
SC
-0.010 (0.022); p= 0.6510
M AN U
KOOS Sport/Rec
-
TE D
KOOS QOL
0.39 (0.15); p=0.0119
EP
KOOS ADL
0.41 (0.14); p=0.0048
Adjusted for BMI, age, sex
-
AC C
KOOS Symptoms
0.45 (0.15); p=0.0040
Reduced model 2
Estimate: slope (SE); p-value. Abbreviations: DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and semimembranosus muscles. ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a plateau or washout enhancement pattern; Nvoxel%, the proportion of voxels reaching either a plateau or washout enhancement pattern relative to the total number of voxels within the muscle VOI; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation, KL, Kellgren-Lawrence. *Data from 93 patients.
Figure 1 MANUSCRIPT Maps of DCE-MRI perfusion variablesACCEPTED in a representative patient with KOA. Axial dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) gradient echo (GRE) T1 w VIBE sequence approximately at the level of basis patella in a representative patient with KOA (A-D). A, The drawing of region of interests (ROIs) around of all the peri-articular knee muscles avoiding major
RI PT
vascular branches. B, Contrast Uptake Pattern: The voxels were automatically assigned and colour coded in a Gadolinium (Gd)-enhancement map to one of four statistically approximated contrast enhancement patterns: washout
SC
(red), persistent (blue), plateau (green), or no enhancement (no color). C, maps of Initial Rate of Enhancement (IRE), and D, maps of Maximum Enhancement (ME). Values of IRE and ME are colour-coded
M AN U
and superimposed on the grey scale dynamic images. The highest values are shown in bright yellow to white and lower values in a spectrum of red colours.
AC C
EP
TE D
White arrows: The popliteal artery. Blue arrows: Enhancement in the synovium in the retro-patellar space.
D
AC C
EP
TE D
M AN U
SC
C
RI PT
ACCEPTED MANUSCRIPT B
A
ACCEPTED MANUSCRIPT
Patients recruited from the LIGHT study at the 1-year follow up (n=134,100%)
RI PT
Figure 2. Flowchart
M AN U
DCE-MRIs obtained (n=97, 72%)
SC
Patients who failed to complete the DCE-MRI (n=37, 28%)
Technical difficulties with the DCE-MRI preventing image analysis (n=3, 2%)
AC C
EP
TE D
DCE-MRIs analysed (n=94, 70%)
ACCEPTED MANUSCRIPT
Supplementary Table Demographics, clinical, and imaging variables of the group that completed the MRI and the group without a completed or legible MRI.
Mean ± SD (min-max)
EP
37 (93 %) 65.2 ± 5.4 (54.7-78.7) 165.3 ± 8.5 (151.0-191.0) 99.1 ± 16.2 (55.1-141.4) 36.3 ± 5.7 (20.5-47.8)
SC
77 (82 %) 65.3 ± 6.5 (52.4-77.4) 165.9 ± 8.1 (148.0-189.0) 88.9 ± 12.6 (68.3-130.2) 32.3 ± 3.7 (25.2-44.2)
63.8 ± 20.8 (16.7-100.0)¤ 64.4 ± 20.0 (17.9-96.4)¤ 68.3 ± 19.1 (17.7-94.1)¤ 45.2 ± 18.4 (6.3-75.0)¤ 25.4 ± 21.4 (0.0-80.0)¤
2.5 ± 0.8 (1-4)*
3.0 ± 1.0 (1-4)
0 (0 %) 10 (10.75 %) 36 (38.7 %) 38 (40.86 %) 9 (9.67 %)
0 (0 %) 2 (5 %) 13 (32.5 %) 9 (22.5 %) 16 (40 %)
M AN U
62.9 ± 17.4 (22.2-100.0) 64.3 ± 19.0 (10.7-100.0) 67.6 ± 17.2 (17.6-98.5) 45.0 ± 18.9 (6.3-93.8) 30.7 ± 23.1 (0.0-95.0)
TE D
Demographics and clinical characteristics Female, no (%) Age (years) Height (cm) Weight (kg) BMI (kg/m2) KOOS scores Pain Symptoms ADL QOL Sport/Rec X-ray KL score KL score, no. (0-4) 0, no. (%) 1, no. (%) 2, no. (%) 3, no. (%) 4, no. (%)
MRI not completed or legible (n=40) Mean ± SD (min-max)
RI PT
MRI completed (n=94)
AC C
Abbreviations: MRI, Magnetic resonance imaging; BMI, Body mass index; KOOS, Knee injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related Quality of life; Sport/Rec, Function in sport and recreation; KL, Kellgren-Lawrence. ¤: Data from 39 patients. *: Data from 93 patients.
S1063-4584(15)01190-5
DOI:
10.1016/j.joca.2015.05.032
Reference:
YJOCA 3508
To appear in:
Osteoarthritis and Cartilage
Received Date: 1 December 2014 Revised Date:
30 April 2015
Accepted Date: 26 May 2015
Please cite this article as: Bandak E, Boesen M, Bliddal H, Riis RG, Gudbergsen H, Henriksen M, Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study, Osteoarthritis and Cartilage (2015), doi: 10.1016/j.joca.2015.05.032. 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.
ACCEPTED MANUSCRIPT 1
Associations between muscle perfusion and symptoms in
2
knee osteoarthritis: a cross sectional study
RI PT
3 4
1
Elisabeth Bandak, PT, MSc, 1,2Mikael Boesen, MD, PhD, 1Henning Bliddal,MD, DMSc, 1,2Robert GC
5
Riis, MD, 1 Henrik Gudbergsen, MD, PhD, 1Marius Henriksen, PT, PhD
7
1
8
Frederiksberg, Copenhagen, Denmark
9
2
The Parker Institute, Dept. of Rheumatology, Copenhagen University Hospital, Bispebjerg &
M AN U
10
SC
6
Department of Radiology, Copenhagen University Hospital, Bispebjerg & Frederiksberg,
Copenhagen, Denmark
11
Email addresses
13
EB: [email protected]
14
MB: [email protected]
15
HB: [email protected]
16
RR: [email protected]
17
HG: [email protected]
18
MH: [email protected]
EP
AC C
19
TE D
12
20 21 22 1
ACCEPTED MANUSCRIPT Correspondence
24
Elisabeth Bandak
25
The Parker Institute
26
Bispebjerg & Frederiksberg Hospitals
27
Nordre Fasanvej 57
28
2000 Frederiksberg
29
Phone: +45 38164169
30
Fax: +45 38164159
31
E-mail: [email protected]
34
Abstract: 248
35
Manuscript text: 2732
36
Tables and figures: 5
37
References: 50
39 40 41
Running title: Muscle perfusion in knee osteoarthritis
AC C
38
M AN U
Word count
EP
33
TE D
32
SC
RI PT
23
42 43 44 2
ACCEPTED MANUSCRIPT Abstract
46
Objective: To investigate the association between muscle perfusion in the peri-articular knee
47
muscles assessed by dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) and
48
symptoms in patients with knee osteoarthritis (KOA).
49
Design: In a cross-sectional setting, muscle perfusion was quantified by DCE-MRI in KOA. Regions
50
of interest were drawn around the peri-articular muscles, summed and averaged into one single
51
“Total Muscle Volume” volume of interest (VOI). Symptoms were assessed via the Knee injury and
52
Osteoarthritis Outcome Score (KOOS) (0: worst; 100: best).
53
Results: DCE-MRI and clinical data were analyzed in 94 patients. The typical participant was a
54
woman with a mean age of 65 years, and a body mass index of 32 kg/m2. Reduced multiple
55
regression models analyzing the association between KOOS and DCE-MRI perfusion variables of
56
Total Muscle Volume showed a statistically significant association between Nvoxel% and KOOS pain
57
(0.41 (SE 0.14); p=0.0048). Nvoxel% was defined as the proportion of highly perfused voxels; i.e. the
58
voxels that show an early and rapid increase on the signal intensity versus time curves, reach a
59
plateau state (plateau pattern) and then showing a relatively rapid decline (washout pattern)
60
relative to the total number of voxels within the muscle VOI.
61
Conclusions: More widespread perfusion in the peri-articular knee muscles was associated with
62
less pain in patients with KOA. These results give rise to investigations of the effects of exercise on
63
muscle perfusion and its possible mediating role in the causal pathway between exercise and pain
64
improvements in the conservative management of KOA.
65
Keywords: Muscle perfusion; Magnetic resonance imaging, knee; osteoarthritis, symptoms
AC C
EP
TE D
M AN U
SC
RI PT
45
3
ACCEPTED MANUSCRIPT 66 67
Introduction Knee osteoarthritis (KOA) is considered a whole-joint disease, including affection of periarticular structures such as muscles and bursa1. Pain is the cardinal symptom in KOA2-4. The
69
pathology of pain has been a main target in understanding the disease with bone marrow lesions
70
(BMLs), synovitis4-6, and the infrapatellar fat pad (IPFP) suggested to play important roles in the
71
generation of pain in KOA7,8. KOA pain is predominantly localized to the joint lines, but patients
72
may also experience a more generalized and unspecific pattern of knee pain involving the whole
73
knee region including peri-articular structures such as muscles9-12.
SC
It has become evident that inflammation plays an important role in the pathogenesis of
M AN U
74
RI PT
68
osteoarthritis4,13,14, and both synovitis and BMLs have shown to be associated with pain and
76
structural progression in KOA15,16. The role of muscle tissue in the symptomatology of KOA is not
77
clear, but findings of inflammatory markers in peri-articular knee muscles in patients with KOA
78
suggest that inflammatory processes are not only confined to articular structures, and that
79
inflammatory processes in peri-articular muscles may contribute to the development of pain17,18.
80
In current practice, conventional radiographs (CR) remain the central component in
81
diagnosing KOA and assess structural progression19 although there is a poor association between
82
radiographic findings and clinical features19,20. This may be due to the fact that CR cannot capture
83
key elements of KOA pathology including inflammation and soft tissue pathology19-21. In contrast
84
to CRs, MRI provides a unique visualization of all the anatomical structures involved in KOA
85
including bone marrow and musculature21,22. The use of dynamic MRI sequences recorded prior to
86
and during intravenous (IV) bolus infusion of a Gadolinium (Gd) contrast medium (Dynamic
87
Contrast Enhanced MRI (DCE-MRI)23 can be used to extract perfusion parameters in both hand and
88
knee joints where the speed, degree, and distribution of the contrast uptake using time-intensity
AC C
EP
TE D
75
4
ACCEPTED MANUSCRIPT curves reflects the histological degree of inflammation in the tissue24 and the change in
90
enhancement over time can be used to assess treatment response in RA2526. DCE-MRI is a valid
91
method to describe musculoskeletal tumors, where high perfusion parameters correlate to higher
92
degree of histologically proven high-grade malignant tumor tissue 27. In KOA, DCE-MRI has only
93
been used in a few studies8,28, but associations between synovial perfusion and synovitis grading
94
assessed both macro- and microscopically have been described28, and recently higher perfusion of
95
the IPFP was associated with pain severity and function88.
SC
96
RI PT
89
When analyzing DCE-MRI images in KOA patients, we have observed that tissue enhancement (i.e. tissues with contrast agent uptake) is not confined to intra-articular structures
98
but also includes different patterns of contrast enhancement in the peri-articular knee muscle
99
tissue as well. It is unknown if the degree of enhancement in these muscles is pathological and
M AN U
97
contributes to the symptoms in KOA or if the contrast uptake seen is due to the basis perfusion of
101
skeletal muscle tissue29-31.
102
TE D
100
The purpose of this study was to investigate the association between muscle perfusion in the peri-articular knee muscles assessed by DCE-MRI and symptoms in patients with
104
KOA. We hypothesized that higher contrast enhancement, and thus perfusion in the peri-articular
105
muscles, would be associated with worse symptoms in patients with KOA. Understanding the role
106
of perfusion in the peri-articular knee muscles in KOA symptomatology may contribute to our
107
understanding of how exercise exerts its impact on symptomatic relief in KOA, which can be used
108
to optimize recommended exercise therapies.
AC C
EP
103
109
5
ACCEPTED MANUSCRIPT 110
Method
111
Study population Cross sectional data from a weight-loss maintenance study in obese patients with
113
radiographically verified tibiofemoral KOA were included in this study (the LIGHT study;
114
ClinicalTrials.gov: NCT00938808). Data used in the present study were gathered at the 1 year
115
follow-up. The LIGHT study was approved by the local health research ethical committee and
116
conducted in accordance with the Helsinki Declaration, as revised in 2000. Written and oral
117
informed consent was obtained from each patient. All participants had previously participated in a
118
68 week weight-loss study; the CAROT study (ClinicalTrials.gov identifier: NCT00655941). The main
119
inclusion criteria were age > 50 years, a body mass index (BMI) ≥ 30 kg/m2, and clinical KOA
120
diagnosis verified by radiographs obtained at baseline of the LIGHT study32. At inclusion, the
121
participants chose a target knee, defined as the most symptomatic knee, for all subsequent
122
assessments.
SC
M AN U
TE D
124
Dynamic Contrast Enhanced MRI protocol (DCE-MRI)
EP
123
RI PT
112
MRI of the target knee, i.e. the most symptomatic knee, was performed on a 3T Siemens
126
Verio® system. The patients were scanned in supine position using a dedicated 16-channel
127
send/receive knee coil. The MRI protocol used has been described in detail earlier8. As
128
recommended by the Danish Health and Medicines Authority, decreased renal function with an
129
estimated glomerular filtration rate (eGFR <60 ml/min/1.73m2) contraindicated administration of
130
IV contrast medium, and thereby DCE-MRI. The DCE-MRI sequence was always performed a
131
minimum of 25 minutes into the scan assuming a relaxed state of the patient with normalization
132
of potential perfusion changes induced by movement.
AC C
125
6
ACCEPTED MANUSCRIPT 133
Image analysis The DCE-MRI sequence was performed in the axial plane in all patients, covering the
135
suprapatellar recess to the insertion of the patellar ligament on the tibia. All DCE-MRI analyses
136
were performed by an investigator blinded to the clinical data (EB) supervised by MB and RR. DCE-
137
MRI slices were analyzed using the computer software Dynamika® version 2.4.4.1 (Image Analysis
138
LTD, London, http://www.imageanalysis.org.uk)33,34. Motion correction between temporal slices
139
was applied on all the available axial DCE-MRI slices35 before regions of interest (ROIs) were drawn
140
around each of the following muscles: vastus lateralis, vastus medialis, biceps femoris, sartorius,
141
gracilis, triceps surae, popliteus, semitendinosus and semimembranosus. Major vascular branches
142
were avoided when drawing the ROIs (Figure 1, A). All the drawn ROIs were summed and averaged
143
into one single “Total Muscle Volume” VOI. Osirix® v. 5.7.1 was used to confirm the anatomical
144
boundaries of the muscles.
SC
M AN U
As previously described in detail 34 ; signal intensity versus time curves can be generated
TE D
145
RI PT
134
for each voxel within a ROI by using appropriate software. Dynamika® uses a robust classification
147
scheme34 where each signal intensity versus time curves automatically are assigned to one of the
148
following enhancement patterns: Tissues with high perfusion are likely to show an early and rapid
149
increase on the signal intensity versus time curves, reach a plateau state (plateau pattern) and
150
then potentially show a relatively rapid decline (washout pattern), whereas tissues with lower
151
perfusion showing a slower increase and potentially do not reach a plateau state (persistent
152
pattern), and tissue with no contrast uptake (no enhancement pattern)25,34. The voxels are
153
automatically assigned and color coded in a Gadolinium (Gd)-enhancement map to one of the four
154
enhancement patterns: washout (red), persistent (blue), plateau (green), or no enhancement (no
155
color)34 26(Figure 1, B).
AC C
EP
146
7
ACCEPTED MANUSCRIPT 156
Various DCE-MRI perfusion variables can be extracted from the signal intensity versus time curves, the most relevant in the setting of musculoskeletal imaging being: i) maximal
158
enhancement (ME), calculated as the maximum signal intensity relative to the baseline signal
159
intensity; ii) the Initial Rate of Enhancement (IRE) i.e. the speed of enhancement or upslope on the
160
signal intensity versus time curve calculated as the relative increase in signal intensity per
161
second(%/s) from time of enhancement onset (in seconds after the contrast injection) until ME is
162
reached, and iii) Nvoxel, the sum of voxels with plateau or washout enhancement patterns, i.e. the
163
highest perfused voxels 24,34,36.
SC
RI PT
157
Values and maps of IRE and ME were generated, color-coded and superimposed on the
165
grey scale dynamic images. The highest values for IRE and ME were shown in bright yellow and
166
lower values in a spectrum of red colors (Figure 1, C-D). Intra- and inter-observer reliability
167
analysis for the Total Muscle Volume (cm3) was performed on a random, repeated subsample
168
(n=10) with a minimum of 4 weeks between drawings using single measures intraclass correlation
169
coefficients (ICC) as a measure of relative reliability and measurement errors (ME) calculated as
170
the square root of the residual mean square values obtained from analyses of variance, as a
171
measure of absolute reliability. The intra-observer reliability of the total muscle volume (cm3) was
172
ICC=0.99 and ME= 6.5 cm3; inter-observer reliability: ICC=0.96 and ME=10.8 cm3. The intra-
173
observer reliability of the IRExNvoxel was ICC=0.97 and ME= 23.3; inter-observer reliability: ICC=0.71
174
and ME=65.0. The intra-observer reliability of the MExNvoxel was ICC=0.99 and ME= 2613.6; inter-
175
observer reliability: ICC=0.91 and ME=7312.1. The intra-observer reliability of the Nvoxel% was
176
ICC=0.99 and ME=1.0%; inter-observer reliability: ICC=0.94 and ME=5.6%.
AC C
EP
TE D
M AN U
164
8
ACCEPTED MANUSCRIPT 177 178 179
Outcome measures The radiographs were scored by a trained radiologist (MB) according to the KellgrenLawrence (KL) grading of radiographic disease severity37. To assess the patients’ symptoms we applied the Knee injury and Osteoarthritis Outcome
181
Score (KOOS)38 questionnaire. KOOS is a self-administered patient-reported outcome measure,
182
assessing five domains of importance to patients with KOA (Pain, Symptoms, Function in daily
183
living, Function in sport and recreation, and Knee related quality of life). Each domain includes
184
individual items answered on 0-4 Likert scale. A normalized 0-100 score (0 indicating the worst,
185
100 indicating the best) is calculated for all domains38. The items in the KOOS questionnaire relate
186
to target knee symptoms during the previous week. In this study we used a Danish touch screen
187
version of KOOS39.
SC
M AN U
188
RI PT
180
The DCE-MRI perfusion variables of the Total Muscle Volume VOI were as follows: IRExNvoxel, MExNvoxel, and Nvoxel%, where IRE and ME were quantified as average measures of the
190
whole segmented region. By multiplying Nvoxel (the sum of voxels with plateau or washout
191
enhancement patterns, i.e. the highest perfused voxels) with the means of the IRE and ME the
192
variables become composite parameters reflecting both the volume and the degree of higher
193
perfused voxels (IRExNvoxel, MExNvoxel)24. Nvoxel% is the proportion of highly perfused voxels
194
reaching either a plateau or washout enhancement patterns relative to the total number of voxels
195
within the muscle VOI. Nvoxel, IRExNvoxel, and MExNvoxel are all variables obtained from DCE-MRI that
196
have been used in previous studies8,24,26,36.
AC C
EP
TE D
189
197
9
ACCEPTED MANUSCRIPT 198
Statistical methods The analyses were pre-specified and conducted on all participants with valid DCE-MRI data,
200
i.e. no imputation for missing data (Figure 2, Flowchart). Five multiple regression models were
201
built to determine the association between the 5 dependent variables (the 5 KOOS subscales and
202
KL score), each with the 3 DCE-MRI perfusion variables of the Total Muscle Volume: IRExNvoxel,
203
MExNvoxel, and Nvoxel% as independent variables. Reductions of the models were done manually
204
eliminating
205
coefficients/slopes) one at a time based on p-values (highest eliminated first). Reductions were
206
only done in regression models that were overall statistically significant (p<0.05). The reduced
207
models were repeated adjusting for possible confounders pre-defined as sex, age, and BMI. To
208
assess the adequacy of the regression model(s) describing the observed data we investigated the
209
model features via the predicted values and the studentized residuals; i.e. the residuals had to be
210
normally distributed (around zero), and be independent of the predicted values. The analyses
211
were done using SAS software, version 9.3 (SAS Inc., Cary, NC, USA). Statistical significance was
212
accepted at p<0.05.
non-significant
(p>0.05)
variables
(i.e.
non-significant
beta-
EP
213
TE D
M AN U
SC
statistically
RI PT
199
Results
215
Patient characteristics
AC C
214
216
In this cross-sectional study, all 134 patients completed the 1-year visit in the LIGHT study.
217
Of these, MRI was performed with 97 patients. Reasons for missing MRI assessment were due to
218
either: (1) the target knee was above 53 cm in circumference and could therefore not fit into the
219
send/receive knee coil (2) decreased renal function with an eGFR < 60 ml/min/1.73m2. Technical
220
difficulties with the DCE-MRI analysis occurred in 3 participants. Thereby the number of analyzed 10
ACCEPTED MANUSCRIPT DCE-MRI data sets was 94 (70% of the study population) (Figure 2). Patient characteristics are
222
available in Table 1. The typical patient participating in this study was a woman aged 65 years
223
with a BMI of 32 kg/m2. The 94 patients who were analyzed in this study had a lower weight and
224
BMI compared with the 40 patients without completed or legible MRIs (Supplementary file).
225
Outcomes
RI PT
221
Summaries of the regression analyses are provided in Table 2. The reduced multiple
227
regression models analyzing the association between KOOS pain and the DCE-MRI perfusion
228
variables of the Total Muscle Volume showed a statistically significant association between Nvoxel%
229
and KOOS pain (Table 2). Adjustment for sex, age, and BMI did not change the results significantly.
230
There were no statistically significant associations between the other KOOS subdomains
231
(Symptoms, Function in daily living, Function in sport and recreation, and Knee related quality of
232
life),the KL score and the DCE-MRI muscle perfusion variables.
233 234
Discussion
TE D
M AN U
SC
226
In this cross sectional study we found a significant association between pain and the
236
proportion of the peri-articular knee muscle tissue with a distinct perfusion pattern among
237
patients with KOA assessed by DCE-MRI. The association suggests that for each percentage point
238
increment in Nvoxel%, the KOOS pain score is 0.41 higher (less pain) in patients with KOA. Thus,
239
individuals with a larger proportion of the peri-articular muscles showing a plateau or washout
240
contrast uptake pattern seemed to have less pain. By consequence, individuals with larger areas
241
with no enhancement or persistent contrast uptake patterns had more pain. IRExNvoxel and
242
MExNvoxel were not associated with pain. We found no associations between the other KOOS
243
subdomains (Symptoms, Function in daily living, Function in sport and recreation, and Knee
AC C
EP
235
11
ACCEPTED MANUSCRIPT 244
related quality of life), the KL score, and the DCE-MRI muscle perfusion variables in the peri-
245
articular knee muscles. These results are in contrast to our hypothesis and are surprising. To the best of our
247
knowledge this is the first study to assess muscle perfusion in patients with KOA, so direct
248
comparisons with other studies are not possible. However, our results are indirectly supported by
249
the fact that muscle weakness is a typical feature of KOA40, and that exercise has beneficial effects
250
on pain in KOA41. Further, exercise has been shown to increase muscle perfusion in both young
251
and elderly healthy subjects42,43, and it is therefore possible that the variation in muscle perfusion
252
in our study reflects variations in the physical fitness status of the cohort. In further (indirect)
253
support of this notion, previous results of the effects of weight loss in this cohort showed that
254
increased muscle strength was associated with decreased pain44.
M AN U
SC
RI PT
246
Our findings are of potential importance in explaining the underlying analgesic mechanisms
256
of exercise interventions in KOA. Several studies have assessed possible theories, including neural
257
pain processing45, biomechanical40,46 and anti-inflammatory mechanisms8,47,48. As the muscles are
258
the tissue primarily affected by exercise, it is relevant to investigate if muscle perfusion is
259
associated with pain in KOA. Future longitudinal studies are wanted, and should investigate if
260
changes in muscle perfusion are a contributing factor explaining the beneficial outcomes of
261
exercise, and if specific types of exercise have specific effects on the muscle perfusion.
AC C
EP
TE D
255
262
Our results indicate that previous findings of a relationship between high pain and high
263
perfusion of the synovium and the IPFP in KOA patients may not be accompanied by similar
264
changes in peri-articular structures8,28. Inflammatory substances have been detected in the
265
quadriceps muscles of individuals with KOA17,28,49. However, these were not associated with pain.
266
Hyperperfusion in articular structures is generally recognized as a sign of inflammation24-26, and as 12
ACCEPTED MANUSCRIPT 267
such undesirable and a potential treatment target in KOA. In contrast, the present results indicate
268
that higher muscle perfusion may be beneficial in KOA and could possibly be a parallel to the
269
changes in muscle capillarization following exercise as observed in other studies43,50. This study has some limitations associated with our cross-sectional design, which
271
precludes an evaluation of a possible relationship between changes in symptoms and DCE-MRI
272
perfusion variables. Also, as the underlying MRI protocol focused on the knee joint, the images
273
were limited to the area of the send/receive knee coil, therefore we were only able to analyze
274
certain parts of the muscles. Further, the likelihood of a false positive finding is not negligible due
275
to the multiple statistical tests. However, the results seem robust even after adjustment for
276
covariates. With a Bonferroni adjustment for multiple tests (6 regressions) the threshold of
277
statistical significance is 0.0083. The statistically significant regression results (KOOS pain; Table 2)
278
respect this threshold for significance, except the covariate adjusted model. Important strengths
279
to our study include the relative large study sample, the use of a 3T MRI system, and the pre-
280
specified analysis plan.
282
Conclusion
EP
281
TE D
M AN U
SC
RI PT
270
In conclusion, this study showed, in a cross-sectional setting, that more widespread
284
perfused voxels in the peri-articular knee muscles was associated with less pain in patients with
285
KOA. These results give rise to investigations of the effects of exercise on muscle perfusion and its
286
possible mediating role in the causal pathway between exercise and pain improvements in the
287
conservative management of KOA.
AC C
283
13
ACCEPTED MANUSCRIPT 288
290 291
Acknowledgements We would like to thank the staff at the Department of Radiology, Bispebjerg and Frederiksberg Hospitals for helping in obtaining the MRI.
RI PT
289
292 293
Financial support
This study was supported by grants from The Danish Physiotherapy Association, The
295
Capital Region of Copenhagen, Bispebjerg-and Frederiksberg Hospitals, The Danish Rheumatism
296
Association, The Oak Foundation, The Velux Foundation, The Cambridge Weight plan UK, The
297
Augustinus Foundation, The A.P.Møller Foundation for the Advancement of Medical Science,
298
Hørslev Fonden, Bjarne Jensens Fond and Aase og Ejnar Danielsens fond.
300
M AN U
The funding sources did not influence study design, collection, analysis and interpretation of data, writing, or the decision to submit the manuscript for publication.
301
304 305 306 307 308
EP
303
Author contributions
EB contributed to the study design, statistical design and analysis, interpretation of the data, drafted and revised the manuscript, and approved the final version.
AC C
302
TE D
299
SC
294
MB contributed to the study design, radiological supervision, and interpretation of the data, revised the manuscript, and approved the final version. HB contributed to the study design, interpretation of the data, revised the manuscript, and approved the final version.
14
ACCEPTED MANUSCRIPT
311 312 313 314 315 316
interpretation of the data, revised the manuscript, and approved the final version. HG contributed to interpretation of the data, revised the manuscript, and approved the final version. MH contributed to the study design, statistical design and analysis, interpretation of the data, revised the manuscript, and approved the final version.
All the authors had access to the data and analyses. EB takes responsibility for the integrity of the work as a whole, from inception to finished article.
M AN U
317 318
RI PT
310
RR contributed in the inter-observer reliability analysis, radiological supervision, and
SC
309
Conflicts of interest
MB reports grants from The Oak Foundation, grants from Abbvie A/S Denmark, non-
320
financial support from ESAOTE MRI vendor, Genoa Italy, non-financial support from Image Analysis
321
LTD London UK, outside the submitted work; All other authors declare no conflicts of interest.
EP AC C
322
TE D
319
15
ACCEPTED MANUSCRIPT
324 325
326 327
References (1) Loeser R, Goldring S, Scanzello C, Goldring MB. Osteoarthritis: a disease of the joint as an organ. Arthritis & Rheumatism 2012;64(6):1697-707.
(2) Schaible HG. Mechanisms of chronic pain in osteoarthritis. Curr Rheumatol Rep 2012
RI PT
323
Dec;14(6):549-56.
(3) Felson DT. The sources of pain in knee osteoarthritis. Curr Opin Rheumatol 2005 Sep;17(5):624-8.
329
(4) Bijlsma JWJ. Osteoarthritis: an update with relevance for clinical practice. Lancet
331
2011;377(9783):2115-26.
M AN U
330
SC
328
(5) Zhang Y, Nevitt M, Niu J, Lewis C, Torner J, Guermazi A, et al. Fluctuation of knee pain and changes
332
in bone marrow lesions, effusions, and synovitis on magnetic resonance imaging. Arthritis Rheum
333
2011 Mar;63(3):691-9.
(6) Guermazi A, Roemer FW, Hayashi D, Crema MD, Niu J, Zhang Y, et al. Assessment of synovitis with
TE D
334
contrast-enhanced MRI using a whole-joint semiquantitative scoring system in people with, or at
336
high risk of, knee osteoarthritis: the MOST study. Ann Rheum Dis 2011 May;70(5):805-11.
337
(7) Crema MD, Felson DT, Roemer FW, Niu J, Marra MD, Zhang Y, et al. Peripatellar synovitis:
EP
335
comparison between non-contrast-enhanced and contrast-enhanced MRI and association with
339
pain. The MOST study. Osteoarthritis Cartilage 2013 Mar;21(3):413-8.
340
AC C
338
(8) Ballegaard C, Riis RG, Bliddal H, Christensen R, Henriksen M, Bartels EM, et al. Knee pain and
341
inflammation in the infrapatellar fat pad estimated by conventional and dynamic contrast-
342
enhanced magnetic resonance imaging in obese patients with osteoarthritis: A cross-sectional
343
study. Osteoarthritis and Cartilage(0). 16
ACCEPTED MANUSCRIPT 344
(9) Thompson LR, Boudreau R, Hannon MJ, Newman AB, Chu CR, Jansen M, et al. The knee pain map:
345
Reliability of a method to identify knee pain location and pattern. Arthritis & Rheumatism 2009 Jun
346
15;61(6):725-31.
350
351
RI PT
349
osteoarthritis of the knee. Osteoarthritis and Cartilage 1998 Sep;6(5):318-23.
(11) McLean SM, Burton M, Bradley L, Littlewood C. Interventions for enhancing adherence with physiotherapy: A systematic review. Manual Therapy 2010 Dec;15(6):514-21.
SC
348
(10) Creamer P, Lethbridge-Cejku M, Hochberg MC. Where does it hurt? Pain localization in
(12) Wood LRJ, Peat G, Thomas E, Duncan R. Knee osteoarthritis in community-dwelling older adults:
M AN U
347
352
are there characteristic patterns of pain location? Osteoarthritis and Cartilage 2007 Jun;15(6):615-
353
23.
(13) Goldring M. Inflammation in osteoarthritis. Current opinion in rheumatology 2011;23(5):471-8.
355
(14) Berenbaum F. Osteoarthritis as an inflammatory disease (osteoarthritis is not osteoarthrosis!).
357
Osteoarthritis and Cartilage 2013 Jan;21(1):16-21.
(15) Hill C, Hunter D, Clancy M, Guermazi A, Genant H, Gale D, et al. Synovitis detected on magnetic
EP
356
TE D
354
resonance imaging and its relation to pain and cartilage loss in knee osteoarthritis. Annals of the
359
Rheumatic Diseases 2007;66(12):1599-603.
360
AC C
358
(16) Tanamas SK, Wluka AE, Pelletier JP, Pelletier JM, Abram F, Berry PA, et al. Bone marrow lesions in
361
people with knee osteoarthritis predict progression of disease and joint replacement: a longitudinal
362
study. Rheumatology (Oxford) 2010 Dec;49(12):2413-9.
17
ACCEPTED MANUSCRIPT 363
(17) Levinger I, Levinger P, Trenerry M, Feller JA, Bartlett JR, Bergman N, et al. Increased inflammatory
364
cytokine expression in the vastus lateralis of patients with knee osteoarthritis. Arthritis &
365
Rheumatism 2011;63(5):1343-8.
(18) Oliveria SA, Felson DT, Cirillo PA, Reed JI, Walker AM. Body weight, body mass index, and incident
367
symptomatic osteoarthritis of the hand, hip, and knee. Epidemiology 1999 Mar;10(2):161-6.
368
(19) Bliddal H, Boesen M, Christensen R, Kubassova O, Torp-Pedersen S. Imaging as a follow-up tool in
371
SC
370
clinical trials and clinical practice. Best Pract Res Clin Rheumatol 2008 Dec;22(6):1109-26.
(20) Bedson J, Croft PR. The discordance between clinical and radiographic knee osteoarthritis: a
M AN U
369
RI PT
366
systematic search and summary of the literature. BMC Musculoskelet Disord 2008;9:116.
(21) Guermazi A, Roemer FW, Burstein D, Hayashi D. Why radiography should no longer be considered a
373
surrogate outcome measure for longitudinal assessment of cartilage in knee osteoarthritis. Arthritis
374
Res Ther 2011;13(6):247.
TE D
372
(22) Menashe L, Hirko K, Losina E, Kloppenburg M, Zhang W, Li L, et al. The diagnostic performance of
376
MRI in osteoarthritis: a systematic review and meta-analysis. Osteoarthritis Cartilage 2012
377
Jan;20(1):13-21.
(23) Verstraete KL, Van der Woude HJ, Hogendoorn PC, De-Deene Y, Kunnen M, Bloem JL. Dynamic
AC C
378
EP
375
379
contrast-enhanced MR imaging of musculoskeletal tumors: basic principles and clinical applications.
380
J Magn Reson Imaging 1996 Mar;6(2):311-21.
381
(24) Axelsen MB, Poggenborg RP, Stoltenberg M, Kubassova O, Boesen M, Horslev-Petersen K, et al.
382
Reliability and responsiveness of dynamic contrast-enhanced magnetic resonance imaging in
383
rheumatoid arthritis. Scand J Rheumatol 2013;42(2):115-22.
18
ACCEPTED MANUSCRIPT 384
(25) Kubassova O, Boesen M, Peloschek P, Langs G, Cimmino MA, Bliddal H, et al. Quantifying Disease
385
Activity and Damage by Imaging in Rheumatoid Arthritis and Osteoarthritis. Annals of the New York
386
Academy of Sciences 2009 Feb 15;1154:207-38.
(26) Boesen M, Kubassova O, Cimmino MA, Ostergaard M, Taylor P, Danneskiold-Samsoe B, et al.
RI PT
387
Dynamic Contrast Enhanced MRI Can Monitor the Very Early Inflammatory Treatment Response
389
upon Intra-Articular Steroid Injection in the Knee Joint: A Case Report with Review of the Literature.
390
Arthritis 2011;2011:578252.
392
393
(27) Nygren AT, Greitz D, Kaijser L. Skeletal muscle perfusion during exercise using Gd-DTPA bolus detection. J Cardiovasc Magn Reson 2000;2(4):263-70.
M AN U
391
SC
388
(28) Loeuille D, Rat AC, Goebel JC, Champigneulle J, Blum A, Netter P, et al. Magnetic resonance imaging in osteoarthritis: which method best reflects synovial membrane inflammation?: Correlations with
395
clinical, macroscopic and microscopic features. Osteoarthritis and Cartilage 2009 Sep;17(9):1186-
396
92.
397
TE D
394
(29) Heinonen I, Saltin B, Kemppainen J, Sipila HT, Oikonen V, Nuutila P, et al. Skeletal muscle blood flow and oxygen uptake at rest and during exercise in humans: a pet study with nitric oxide and
399
cyclooxygenase inhibition. Am J Physiol Heart Circ Physiol 2011 Apr;300(4):H1510-H1517.
400
(30) Levy M. Special Circulations. In: Berne R, Levy M, Koeppen B, Stanton B, editors. Physiology. Fifth
402
AC C
401
EP
398
Edition ed. Missouri: Mosby; 2004. p. 413-32.
(31) Bajwa A, Wesolowski R, Patel A, Saha P, Ludwinski F, Smith A, et al. Assessment of tissue perfusion
403
in the lower limb: current methods and techniques under development. Circ Cardiovasc Imaging
404
2014 Sep;7(5):836-43.
19
ACCEPTED MANUSCRIPT 405
(32) Riecke BF, Christensen R, Christensen P, Leeds AR, Boesen M, Lohmander LS, et al. Comparing two
406
low-energy diets for the treatment of knee osteoarthritis symptoms in obese patients: a pragmatic
407
randomized clinical trial. Osteoarthritis and Cartilage 2010 Jun;18(6):746-54.
(33) Kubassova O, Boesen M, Boyle RD, Cimmino MA, Jensen KE, Bliddal H, et al. Fast and robust
RI PT
408 409
analysis of dynamic contrast enhanced MRI datasets. Med Image Comput Comput Assist Interv
410
2007;10(Pt 2):261-9.
(34) Kubassova O, Boesen M, Cimmino MA, Bliddal H. A computer-aided detection system for
SC
411
rheumatoid arthritis MRI data interpretation and quantification of synovial activity. European
413
Journal of Radiology 2010;74(3):67-72.
414
M AN U
412
(35) Boesen M, Kubassova O, Parodi M, Bliddal H, Innocenti S, Garlaschi G, et al. Comparison of the manual and computer-aided techniques for evaluation of wrist synovitis using dynamic contrast-
416
enhanced MRI on a dedicated scanner. European Journal of Radiology 2011 Feb;77(2):202-6.
417
(36) Boesen M, Kubassova O, Bouert R, Axelsen M, Ostergaard M, Cimmino MA, et al. Correlation
TE D
415
between computer-aided dynamic gadolinium-enhanced MRI assessment of inflammation and
419
semi-quantitative synovitis and bone marrow oedema scores of the wrist in patients with
420
rheumatoid arthritis-a cohort study. Rheumatology 2012;51(1):134-43.
422
423
AC C
421
EP
418
(37) Kelgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Annals of the Rheumatic Diseases 1957;16(4):494-502.
(38) Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD. Knee Injury and Osteoarthritis Outcome
424
Score (KOOS)--development of a self-administered outcome measure. J Orthop Sports Phys Ther
425
1998 Aug;28(2):88-96.
20
ACCEPTED MANUSCRIPT 426
(39) Gudbergsen H, Bartels E, Krusager P, Waehrens E, Christensen R, nneskiold-Samsoe B, et al. Testretest of computerized health status questionnaires frequently used in the monitoring of knee
428
osteoarthritis: a randomized crossover trial. BMC Musculoskeletal Disorders 2011;12(1):190.
429
(40) Bennell KL, Wrigley TV, Hunt MA, Lim BW, Hinman RS. Update on the Role of Muscle in the Genesis
RI PT
427
430
and Management of Knee Osteoarthritis. Rheumatic Disease Clinics of North America 2013
431
Feb;39(1):145-76.
(41) McAlindon TE, Bannuru RR, Sullivan MC, Arden NK, Berenbaum F, Bierma-Zeinstra SM, et al. OARSI
SC
432
guidelines for the non-surgical management of knee osteoarthritis. Osteoarthritis and Cartilage
434
2014 Mar;22(3):363-88.
436
437 438
439
(42) Harris BA. The influence of endurance and resistance exercise on muscle capillarization in the elderly: a review. Acta Physiol Scand 2005 Oct;185(2):89-97.
(43) Hoier B, Hellsten Y. Exercise-induced capillary growth in human skeletal muscle and the dynamics
TE D
435
M AN U
433
of VEGF. Microcirculation 2014 May;21(4):301-14.
(44) Henriksen M, Christensen R, Danneskiold-Samsoe B, Bliddal H. Changes in lower extremity muscle mass and muscle strength after weight loss in obese patients with knee osteoarthritis: a
441
prospective cohort study. Arthritis Rheum 2012 Feb;64(2):438-42.
AC C
442
EP
440
(45) Henriksen M, Klokker L, Graven-Nielsen T, Bartholdy C, Jorgensen TS, Bandak E, et al. Exercise
443
therapy reduces pain sensitivity in patients with knee osteoarthritis: A randomized controlled trial.
444
Arthritis Care Res (Hoboken ) 2014 Jun 6.
21
ACCEPTED MANUSCRIPT 445
(46) Bennell KL, Hunt MA, Wrigley TV, Hunter DJ, McManus FJ, Hodges PW, et al. Hip strengthening
446
reduces symptoms but not knee load in people with medial knee osteoarthritis and varus
447
malalignment: a randomised controlled trial. Osteoarthritis Cartilage 2010 May;18(5):621-8.
(47) Helmark I, Mikkelsen U, Børglum J, Rothe A, Petersen M, Andersen O, et al. Exercise increases
RI PT
448 449
interleukin-10 levels both intraarticularly and peri-synovially in patients with knee osteoarthritis: a
450
randomized controlled trial. Arthritis research & therapy 2010;12(4):R126.
SC
452
(48) Petersen AM, Pedersen BK. The anti-inflammatory effect of exercise. Journal of applied physiology 2005 Apr 1;98(4):1154-62.
M AN U
451
453
(49) Levinger P, Caldow MK, Feller JA, Bartlett JR, Bergman NR, McKenna MJ, et al. Association between
454
skeletal muscle inflammatory markers and walking pattern in people with knee osteoarthritis.
455
Arthritis & Rheumatism 2011;63(12):1715-21.
(50) Weber MA, Hildebrandt W, Schröder L, Kinscherf R, Krix M, Bachert P, et al. Concentric resistance
457
training increases muscle strength without affecting microcirculation. European Journal of
458
Radiology 2010 Mar;73(3):614-21.
461 462 463
EP
460
AC C
459
TE D
456
464
22
ACCEPTED MANUSCRIPT Table and figure legends
466
Figure 1
467
Maps of DCE-MRI perfusion variables in a representative patient with KOA.
468
Axial dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) gradient echo (GRE) T1
469
w VIBE sequence approximately at the level of basis patella in a representative patient with KOA
470
(A-D).
471
A, The drawing of region of interests (ROIs) around of all the peri-articular knee muscles avoiding
472
major vascular branches.
473
B, Contrast Uptake Pattern: The voxels are automatically assigned and color coded in a Gadolinium
474
(Gd)-enhancement map to one of the four enhancement patterns: washout (red), persistent
475
(blue), plateau (green), or no enhancement (no color). C, maps of Initial Rate of Enhancement
476
(IRE), and D, maps of Maximum Enhancement (ME). Values of IRE and ME are color-coded and
477
superimposed on the grey scale dynamic images. The highest values are shown in bright yellow to
478
white and lower values in a spectrum of red colors.
479
White arrows: The popliteal artery. Blue arrows: Enhancement in the synovium in the retro-
480
patellar space.
AC C
EP
TE D
M AN U
SC
RI PT
465
481
482
Figure 2
483
Flowchart
484 23
ACCEPTED MANUSCRIPT Table 1
486
Demographics, self-reported symptoms, and imaging variables from the 94 patients with KOA
487
included in this study.
RI PT
485
488
Abbreviations: BMI, Body mass index; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL,
490
Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and
491
recreation; KL, Kellgren-Lawrence, DCE-MRI, dynamic contrast enhanced magnetic resonance
492
imaging;
493
Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of
494
the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and
495
semimembranosus muscles; ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel,
496
the sum of voxels with a plateau or washout enhancement pattern. Nvoxel%, the proportion of
497
voxels reaching either a plateau or washout enhancement pattern relative to the total number of
498
voxels within the muscle VOI.
499
*: Data from 93 patients.
500 501 502
AC C
EP
TE D
M AN U
SC
489
503
Table 2
504
Multiple regression analysis with KOOS subscales and KL score as dependent variables and DCE-
505
MRI perfusion variables of Total Muscle Volume as independent variables, (n=94).
506 24
ACCEPTED MANUSCRIPT Estimate: slope (SE); p-value.
508
Abbreviations: DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle
509
Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus
510
lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and
511
semimembranosus muscles.
512
ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a
513
plateau or washout enhancement pattern; Nvoxel%, the proportion of voxels reaching either a
514
plateau or washout enhancement pattern relative to the total voxel number of voxels within the
515
muscle VOI; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living;
516
QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation, KL, Kellgren-
517
Lawrence.
518
*Data from 93 patients.
TE D
519
M AN U
SC
RI PT
507
Supplementary Table
521
Demographics, clinical, and imaging variables of the group that completed the MRI and the group
522
without a completed or legible MRI.
AC C
523
EP
520
524
Abbreviations: MRI, Magnetic resonance imaging; BMI, Body mass index; KOOS, Knee injury and
525
Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related Quality of life;
526
Sport/Rec, Function in sport and recreation; KL, Kellgren-Lawrence.
527
¤: Data from 39 patients. *: Data from 93 patients.
528 25
ACCEPTED MANUSCRIPT Table 1. Demographics, self-reported symptoms, and imaging variables from the 94 patients with KOA included in this study.
77 (82 %) 65.3 ± 6.5 (52.4-77.4) 165.9 ± 8.1 (148.0-189.0) 88.9 ± 12.6 (68.3-130.2) 32.3 ± 3.7 (25.2-44.2)
M AN U
SC
62.9 ± 17.4 (22.2-100.0) 64.3 ± 19.0 (10.7-100.0) 67.6 ± 17.2 (17.6-98.5) 45.0 ± 18.9 (6.3-93.8) 30.7 ± 23.1 (0.0-95.0) 2.5 ± 0.8 (1-4) 0 (0 %) 10 (10.75 %) 36 (38.7 %) 38 (40.86 %) 9 (9.67 %)
EP
TE D
Demographics and clinical characteristics Female, no (%) Age (years) Height (cm) Weight (kg) BMI (kg/m2) KOOS scores Pain Symptoms ADL QOL Sport/Rec X-ray KL score* KL score, no. (0-4) 0, no. (%) 1, no. (%) 2, no. (%) 3, no. (%) 4, no. (%) DCE-MRI perfusion variables Total Muscle Volume Mean IRExNvoxel Mean MExNvoxel Nvoxel%
RI PT
Mean ± SD (min-max)
218.8 ± 119.7 (50.3-663.1) 74,172.3 ± 21,061.4 (23,233.7-12,8382.4) 74.9 ± 12.2 (17.6-95.3)
AC C
Abbreviations: BMI, Body mass index; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation; KL, KellgrenLawrence, DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and semimembranosus muscles; ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a plateau or washout enhancement pattern. Nvoxel%, the proportion of voxels reaching either a plateau or washout enhancement pattern relative to the total number of voxels within the muscle VOI. *: Data from 93 patients.
ACCEPTED MANUSCRIPT Table 2 Multiple regression analysis with KOOS subscales and KL score as dependent variables and DCE-MRI perfusion variables of Total Muscle Volume as independent variables, (n=94). Dependent variable
Independent variables
Model
Reduced model 1
KOOS pain
IRExNvoxel
-0.016 (0.019); p=0.4081
-0.011 (0.016); p=0.4691
-
-
MExNvoxel
0.000052 (0.00013); p=0.6805
-
-
-
Nvoxel%
0.41 (0.18), p=0.0258
IRExNvoxel
-0.016 (0.021); p=0.4380
MExNvoxel
0.00016 (0.00014); p=0.2584
Nvoxel%
0.22 (0.20); p=0.2766
IRExNvoxel
-0.020 (0.019); p=0.2908
MExNvoxel
0.00010 (0.00013); p=0.4172
Nvoxel%
0.26 (0.18); p=0.1505
IRExNvoxel
KL score*
RI PT -
-
0.00015 (0.00016); p= 0.3496
-
-
0.22 (0.24); p= 0.3568
-
-
-0.015 (0.02); p= 0.4491
-
-
0.00015 (0.00014); p= 0.2905
-
-
0.15 (0.21); p= 0.4851
-0.014 (0.021); p=0.4997
-
-
-0.008 (0.022); p=0.7168
MExNvoxel
0.00013 (0.00014); p=0.3443
-
-
0.00016 (0.00016); p=0.3170
Nvoxel%
0.11 (0.20); p=0.5826
-
-
0.03 (0.24) p= 0.8884
IRExNvoxel
-0.008 (0.026); p=0.7560
-
-
0.004 (0.027); p= 0.8918
MExNvoxel
0.00018 (0.00017); p=0.2960
-
-
0.00021 (0.00019); p= 0.2756
Nvoxel%
-0.09 (0.25); p=0.7132
-
-
-0.20 (0.29): p= 0.4870
IRExNvoxel
0.000 (0.001); p=0.5032
-
-
0.000 (0,001); p=0.773
MExNvoxel
0.00000 (0.00001); p=0.6399
-
-
0.00000 (0.00001); p=0.9082
Nvoxel%
0.00 (0.01); p=0.7829
-
-
0.00 (0.01); p= 0.6998
SC
-0.010 (0.022); p= 0.6510
M AN U
KOOS Sport/Rec
-
TE D
KOOS QOL
0.39 (0.15); p=0.0119
EP
KOOS ADL
0.41 (0.14); p=0.0048
Adjusted for BMI, age, sex
-
AC C
KOOS Symptoms
0.45 (0.15); p=0.0040
Reduced model 2
Estimate: slope (SE); p-value. Abbreviations: DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; Total Muscle Volume, a volume of interest (VOI) consisting of the summed and averaged ROIs of the vastus lateralis, vastus medialis, biceps femoris, sartorius, gracilis, triceps surae, popliteus, and semimembranosus muscles. ME, Maximal enhancement; IRE, Initial rate of enhancement; Nvoxel, the sum of voxels with a plateau or washout enhancement pattern; Nvoxel%, the proportion of voxels reaching either a plateau or washout enhancement pattern relative to the total number of voxels within the muscle VOI; KOOS, Knee Injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related quality of life; Sport/Rec, Function in sport and recreation, KL, Kellgren-Lawrence. *Data from 93 patients.
Figure 1 MANUSCRIPT Maps of DCE-MRI perfusion variablesACCEPTED in a representative patient with KOA. Axial dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) gradient echo (GRE) T1 w VIBE sequence approximately at the level of basis patella in a representative patient with KOA (A-D). A, The drawing of region of interests (ROIs) around of all the peri-articular knee muscles avoiding major
RI PT
vascular branches. B, Contrast Uptake Pattern: The voxels were automatically assigned and colour coded in a Gadolinium (Gd)-enhancement map to one of four statistically approximated contrast enhancement patterns: washout
SC
(red), persistent (blue), plateau (green), or no enhancement (no color). C, maps of Initial Rate of Enhancement (IRE), and D, maps of Maximum Enhancement (ME). Values of IRE and ME are colour-coded
M AN U
and superimposed on the grey scale dynamic images. The highest values are shown in bright yellow to white and lower values in a spectrum of red colours.
AC C
EP
TE D
White arrows: The popliteal artery. Blue arrows: Enhancement in the synovium in the retro-patellar space.
D
AC C
EP
TE D
M AN U
SC
C
RI PT
ACCEPTED MANUSCRIPT B
A
ACCEPTED MANUSCRIPT
Patients recruited from the LIGHT study at the 1-year follow up (n=134,100%)
RI PT
Figure 2. Flowchart
M AN U
DCE-MRIs obtained (n=97, 72%)
SC
Patients who failed to complete the DCE-MRI (n=37, 28%)
Technical difficulties with the DCE-MRI preventing image analysis (n=3, 2%)
AC C
EP
TE D
DCE-MRIs analysed (n=94, 70%)
ACCEPTED MANUSCRIPT
Supplementary Table Demographics, clinical, and imaging variables of the group that completed the MRI and the group without a completed or legible MRI.
Mean ± SD (min-max)
EP
37 (93 %) 65.2 ± 5.4 (54.7-78.7) 165.3 ± 8.5 (151.0-191.0) 99.1 ± 16.2 (55.1-141.4) 36.3 ± 5.7 (20.5-47.8)
SC
77 (82 %) 65.3 ± 6.5 (52.4-77.4) 165.9 ± 8.1 (148.0-189.0) 88.9 ± 12.6 (68.3-130.2) 32.3 ± 3.7 (25.2-44.2)
63.8 ± 20.8 (16.7-100.0)¤ 64.4 ± 20.0 (17.9-96.4)¤ 68.3 ± 19.1 (17.7-94.1)¤ 45.2 ± 18.4 (6.3-75.0)¤ 25.4 ± 21.4 (0.0-80.0)¤
2.5 ± 0.8 (1-4)*
3.0 ± 1.0 (1-4)
0 (0 %) 10 (10.75 %) 36 (38.7 %) 38 (40.86 %) 9 (9.67 %)
0 (0 %) 2 (5 %) 13 (32.5 %) 9 (22.5 %) 16 (40 %)
M AN U
62.9 ± 17.4 (22.2-100.0) 64.3 ± 19.0 (10.7-100.0) 67.6 ± 17.2 (17.6-98.5) 45.0 ± 18.9 (6.3-93.8) 30.7 ± 23.1 (0.0-95.0)
TE D
Demographics and clinical characteristics Female, no (%) Age (years) Height (cm) Weight (kg) BMI (kg/m2) KOOS scores Pain Symptoms ADL QOL Sport/Rec X-ray KL score KL score, no. (0-4) 0, no. (%) 1, no. (%) 2, no. (%) 3, no. (%) 4, no. (%)
MRI not completed or legible (n=40) Mean ± SD (min-max)
RI PT
MRI completed (n=94)
AC C
Abbreviations: MRI, Magnetic resonance imaging; BMI, Body mass index; KOOS, Knee injury and Osteoarthritis Outcome Score; ADL, Function in daily living; QOL, Knee related Quality of life; Sport/Rec, Function in sport and recreation; KL, Kellgren-Lawrence. ¤: Data from 39 patients. *: Data from 93 patients.