Content uploaded by Laurence A Bradley
Author content
All content in this area was uploaded by Laurence A Bradley on Dec 08, 2015
Content may be subject to copyright.
Content uploaded by Laurence A Bradley
Author content
All content in this area was uploaded by Laurence A Bradley on Dec 08, 2015
Content may be subject to copyright.
Joint inflammation is associated with pain sensitization in knee osteoarthritis: The
Multicenter Osteoarthritis Study
1
Tuhina Neogi, MD, PhD
2
Ali Guermazi, MD
2
Frank Roemer, MD
3
Michael Nevitt, PhD
4
Joachim Scholz, MD
5
Lars Arendt-Nielsen, Dr.Med.Sci., PhD
6
Clifford Woolf, MB, BCh, PhD
1
Jingbo Niu, DSc
7
Laurence A. Bradley, PhD
1
Emily Quinn, BSc
8
Laura Frey Law, MPT, MS, PhD
1
Boston University School of Medicine, Clinical Epidemiology Unit, Boston, MA;
2
Boston
University School of Medicine, Radiology, Boston, MA;
3
UCSF, Epidemiology and Biostatistics,
San Francisco, CA;
4
Columbia University College of Physicians and Surgeons, Anesthesiology, NY,
NY;
5
Aalborg University, Health Science and Technology, Aalborg, Denmark;
6
Boston Children’s
Hospital, FM Kirby Neurobiology Center, Boston, MA;
7
University of Alabama at Birmingham,
Clinical Immunology and Rheumatology, Birmingham, AL;
8
University of Iowa Carver College of
Medicine, Physical Therapy & Rehabilitation Science, Iowa City, IA
Running Title: MRI lesions and sensitization in knee OA
Funding Acknowledgement:
Dr. Neogi: NIH P60 AR47785, R01 AR062506
The MOST Study: NIH U01
AG18820, U01 AG18832, U01 AG18947, and U01AG19079
No support from commercial sources
No financial conflicts of interest
Corresponding Author:
Tuhina Neogi, MD, PhD, FRCPC, 650 Albany Street, Suite X200, Clin Epi Unit, Boston, MA, 02118
Fax: 617-638-5239
Email: tneogi@bu.edu
Full Length
Arthritis & Rheumatology
DOI 10.1002/art.39488
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process which may lead to
differences between this version and the Version of Record. Please cite this article as an
‘Accepted Article’, doi: 10.1002/art.39488
© 2015 American College of Rheumatology
Received: May 29, 2015; Revised: Aug 03, 2015; Accepted: Oct 22, 2015
This article is protected by copyright. All rights reserved.
2
Abstract
Objective: Pain sensitization is associated with pain severity in knee osteoarthritis, but its cause
in humans is not well-understood. We examined whether inflammation, assessed as synovitis
and effusion on MRI, or mechanical load, assessed as bone marrow lesions (BMLs), were
associated with sensitization in knee osteoarthritis.
Methods: Subjects in the Multicenter Osteoarthritis Study, a NIH-funded cohort of persons with
or at risk of knee osteoarthritis, had knee radiographs and MRIs, and standardized quantitative
sensory testing (QST) measures (temporal summation, pressure pain threshold (PPT)) at the
wrist and patellae obtained at baseline and two years later. We examined the relation of
synovitis, effusion, and BMLs to temporal summation and PPT cross-sectionally and
longitudinally.
Results: There were 1111 subjects in the study sample (mean age 67, mean BMI 30, 62%
female). Synovitis was associated with a significant decrease in PPT at the patella (i.e., more
sensitized) over two years (adjusted beta: -0.30, 95% CI -0.52 to -0.08). Effusion was similarly
associated with a decrease in PPT at the wrist (-0.24, 95% CI -0.41 to -0.24) and with risk of
incident temporal summation (adjusted OR 1.54, 95% CI 1.01-2.36). BMLs were not associated
with either QST measure.
Conclusion: Inflammation, as evidenced by synovitis or effusion, is associated with pain
sensitization in knee osteoarthritis. In contrast, BMLs do not appear to contribute to
sensitization in knee osteoarthritis. Early targeting of inflammation is a reasonable strategy to
Page 2 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
3
test for prevention of sensitization and through this, reduction of pain severity in knee
osteoarthritis.
Introduction
Pain remains the primary symptomatic complaint of persons with knee osteoarthritis, the most
common form of arthritis in the United States,
1
yet the determinants of this pain remain poorly
understood. Identification of key factors leading to pain is critical to improving management of
the symptoms of knee osteoarthritis and preventing the emergence of such symptoms.
Alterations in the neurologic processing of nociceptive signaling leading to an enhanced pain
facilitation has been increasingly recognized as one mechanism by which pain in knee
osteoarthritis may become chronic and persistent.
2
Specifically, an increased responsiveness
(sensitization) of peripheral or central nociceptive neurons appears to contribute to the pain
experience in knee osteoarthritis. Sensitization leads to heightened pain sensitivity, thereby
contributing to a more severe pain experience. Pain sensitization, as assessed by quantitative
sensory testing, has been associated with painful knee osteoarthritis when compared with pain-
free controls,
3-10
and with pain severity independent of knee osteoarthritis severity.
11-13
While
such sensitization may be a promising target for ameliorating pain severity in knee
osteoarthritis, the mechanisms by which this sensory sensitization occurs in humans is
incompletely understood. Addressing this knowledge gap would identify a potentially novel
therapeutic strategy for prevention of the typically inevitable progression of pain worsening in
knee osteoarthritis. In animal models, sustained inflammatory stimuli or mechanical tissue
Page 3 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
4
injury can lead to peripheral and central sensitization.
14-22
Whether the same types of stimuli
are important in the development of sensitization in humans is not known, however. There is
some suggestion, though, that sensitization may in fact be influenced by genetic predisposition
with generalized lower pain thresholds that may become manifest once nociceptive input from
the osteoarthritic joint is received.
13
On the other hand, it is possible that only certain pathologic features of osteoarthritis, such as
those that are inflammatory or those that reflect mechanical injury of the bone, which is richly
innervated with nociceptors, lead to sensitization. While knee osteoarthritis is traditionally
considered a systemically non-inflammatory arthritis, contemporary studies using magnetic
imaging resonance (MRI) have demonstrated local inflammation as evidenced by synovitis and
effusion; the latter is also often evident clinically. In contrast, bone marrow lesions (BMLs) are
considered to be predominantly (micro-)traumatic lesions related to excessive mechanical load
or remodeling related to tissue injury. Both the inflammatory lesions (i.e., synovitis and
effusion) and BMLs are the primary pathologic lesions that have been consistently associated
with pain in knee osteoarthritis,
23-25
but the mechanism by which they contribute to pain has
not been elucidated. Given the development of sensitization in animal models related to
inflammatory and/or tissue injury, we sought to determine whether inflammatory and
mechanical lesions in knee osteoarthritis are associated with pain sensitization in humans. If
either of these types of lesions were to lead to sensitization, they would be attractive as early
therapeutic targets to prevent the occurrence of sensitization, with the expectation that this
Page 4 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
5
would in turn prevent the eventual development of chronic and/or more severe pain in knee
osteoarthritis.
Materials and Methods
Study Sample
The Multicenter Osteoarthritis (MOST) Study is a NIH-funded longitudinal cohort of older adults
with or at risk of knee osteoarthritis. At baseline, there were 3026 subjects, aged 50-79, who
were recruited from Birmingham, Alabama and Iowa City, Iowa. Details of the cohort have been
published elsewhere.
26
In brief, subjects were assessed at baseline, 30-months, 60-months, and
84-months with imaging, standardized questionnaires, and objective measures of relevance to
knee osteoarthritis. The institutional review boards at the University of Alabama at
Birmingham, University of Iowa at Iowa City, University of California at San Francisco, and
Boston University Medical Center approved the study protocol.
The current study sample comprised subjects who attended the 60-month and 84-month study
visits; these were the first visits at which standardized quantitative sensory testing, i.e.,
measures of sensitization, were obtained. Eighty-eight subjects who screened positive for
possible peripheral neuropathy
27
were excluded from this analysis. For the purposes of this
study, the 60-month visit will be considered the baseline visit, and the 84-month visit will be
considered the follow-up visit.
MRI Acquisition and Assessment
Page 5 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
6
MRIs of both knees were obtained on a 1.0 Telsa dedicated extremity unit (ONI MSK Extreme
1.0 T, GE Healthcare, Wilmington, MA) with a circumferential extremity coil using fat-
suppressed fast spin-echo intermediate-weighted sequences in two planes, sagittal (TR = 4800
ms, TE = 35 ms, 3 mm slice thickness, 0 mm interslice gap, 32 slices, 288 x 192 matrix, 2
excitations (NEX), 140 x 140 mm field of view (FOV), echo train length (ETL) = 8) and axial (TR =
4680 ms, TE = 13 ms, 3 mm slice thickness, 0 mm interslice gap, 20 slices, 288 x 192 matrix, 2
NEX, 140 x 140 mm FOV, ETL = 8), and a short tau inversion-recovery (STIR) sequence in the
coronal plane (TR = 6650 ms, TE = 15 ms, TI = 100 ms, 3 mm slice thickness, 0 mm interslice gap,
28 slices, 256 x 192 matrix, 2 NEX, 140 mm
2
FOV, ETL = 8). Examinations were performed at the
University of Alabama at Birmingham and at the University of Iowa at Iowa City using the same
MRI unit at baseline and follow-up. Although these are non-contrast-enhanced sequences,
both synovitis and effusion assessments have been validated using this approach,
28-32
and
synovitis on non-contrast-enhanced MRI has been correlated with macroscopic and microscopic
evidence of inflammation.
33
All subjects in MOST who attended the 60- and 84-month study visits had bilateral knee MRIs
obtained unless there were contraindications or endstage OA was present (i.e., Kellgren and
Lawrence grade 4). For each subject, one knee with acceptable quality MRIs at both time-points
was selected for longitudinal reading; if both knees were eligible for readings, one was
randomly selected. The MRIs were read by four experienced musculoskeletal radiologists,
blinded to clinical and radiographic data, using the Whole-Organ Magnetic Resonance Imaging
Score (WORMS).
34
Synovitis was scored 0-3 in the intercondylar region and infrapatellar fat pad
Page 6 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
7
(Hoffa’s fat pad), which represent a surrogate for true synovitis,
29
and is known as Hoffa-
synovitis.
28
Effusion was scored 0-3 in the suprapatellar pouch, often referred to as effusion-
synovitis because it reflects a composite of effusion and synovitis on non-contrast enhanced
MRIs.
28,31
BMLs were scored 0-3 in all 14 WORMS’ subregions of the knee, and excluded tibial
spinous BMLs since this region is not subchondral. Examples of each grade of each MRI lesion
are provided in Figure 1. The ranges of weighted kappas for inter-reader reliability among all
four readers for presence vs. absence of each feature were: 0.80-0.89 for BMLs; 0.16-0.60 for
synovitis; and 0.57-0.86 for effusion.
Quantitative Sensory Testing
Pressure pain threshold (PPT), a measure of sensitivity to pain evoked by mechanical
stimulation of nociceptors,
35-37
can be reliably assessed in knee osteoarthritis with pressure
algometry.
4-7,9,13
We obtained PPTs at baseline and follow-up 24 months later at both patellae
and at the wrist, as published previously.
13
In brief, PPT assessed at a diseased site (e.g., knee
with osteoarthritis) is thought to reflect peripheral sensitization, while when assessed at a
distant, normal site (e.g., wrist), it is thought to reflect central sensitization, or a generalized
level of pain sensitivity. PPT was assessed by applying an algometer (1cm
2
rubber tip; Wagner,
FDIX25, Greenwich, CT) at a rate of 0.5 kg/second as the point at which participants verbally
indicated the pressure first changed to slight pain. The PPT at each anatomic site was calculated
by averaging 3 trials; the position of the algometer was slightly altered with each trial to avoid
sensitization at the test site. Lower PPTs represent a greater degree of sensitization or pain
sensitivity.
Page 7 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
8
Mechanical temporal summation, an augmented response to repetitive mechanical stimulation,
is a sensitive and valid measure of central pain amplification, a feature of central sensitization,
in knee OA.
9,11,13
Mechanical temporal summation was assessed at baseline and follow-up 24
months later using a weighted 60g monofilament (Aalborg University, Denmark) at the patellae
and wrist, as previously described.
13
In brief, subjects provided a numerical pain rating (0-10) to
an initial trial of 4 stimulations, followed by a pain rating at the end of a train of 30 stimulations
applied at a frequency of 1 Hz, and again 15 seconds post-stimulation (“after-sensations”).
Temporal summation was defined as being present when, compared with the initial trial, the
subject reported increased pain following the repeated mechanical stimulation at the site being
tested.
Assessors were blinded to clinical and imaging data. Fourteen-day test-retest reliability for PPT
was 0.85-0.90 (intra-class coefficients) and for temporal summation was 0.61 (kappa).
We refer to these measures as ‘sensitization’ hereafter, but acknowledge that they may also
reflect heightened pain sensitivity.
Potential confounders
Potential confounders included age, sex, body mass index (BMI), race, study site, radiographic
osteoarthritis severity (Kellgren and Lawrence grade for the tibiofemoral joint, and presence of
patellofemoral osteoarthritis), depressive symptoms (Center for Epidemiologic Studies
Page 8 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
9
Depression Scale
38
), catastrophizing (from the Coping Strategies Questionnaire
39
), widespread
pain using a validated standard homunculus,
40
and use of analgesics (NSAIDs, COX-2 inhibitors,
opiates, acetaminophen).
Statistical Analysis
We defined each MRI feature of interest (i.e., synovitis, effusion, and BMLs) as being present if
their WORMS score was ≥1. For BMLs, we additionally evaluated a “BML burden” as a sum of
the BML scores across all of the subregions.
To examine the relation of these MRI features to baseline presence of pain sensitization,
change in pain sensitization (assessed as change in PPT) over two years, and to the new
development of pain sensitization (assessed as development of temporal summation) over two
years, we performed the following. Among all subjects, we evaluated the relation of each MRI
feature at baseline in separate models to baseline PPT and to change in PPT over two years
using linear regression (change in PPT at both the patella and at the wrist was normally
distributed). We evaluated the relation of each of the MRI features at baseline to baseline
presence of temporal summation and to incident temporal summation (i.e., new development
of summation) over two years among subjects who were free of temporal summation at
baseline using logistic regression. In exploratory analyses, we examined the relation of
persistence and resolution of the MRI features compared with their absence at both time-
points to change in PPT over two years using linear regression. To address the same question
for temporal summation, which is a binary outcome, we used a self-matched case-control
Page 9 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
10
study, in which subjects who exhibited temporal summation in one visit, but not the other,
were analyzed using conditional logistic regression.
For all analyses, the MRI features were examined in relation to the quantitative sensory testing
measurement in the ipsilateral patella and at the wrist in separate models. All analyses were
adjusted for the potential confounders listed above. Analyses were conducted using SAS
version 9.3 (SAS Institute, Gary, NC).
Results
Of the 1185 subjects whose knee MRIs were read at baseline and follow-up, 1,111 subjects had
quantitative sensory testing at both the patella and the wrist at baseline and follow-up 24-
months later; these subjects were included in the evaluation of the relation of the MRI feature
to baseline quantitative sensory measurement and to change in PPT. Of these subjects, 716
were eligible for the evaluation of the relation of MRI features to incident temporal summation
(i.e., 716 were free of temporal summation at baseline). The mean age of the whole study
sample was 66.9 years (SD 7.5), with a mean BMI of 29.7 kg/m
2
(SD 4.8); 62% were female
(Table 1). Thirty-eight percent had radiographic knee osteoarthritis at baseline, the mean
WOMAC pain score was 2.2 (SD 2.9), and 21% reported frequent knee pain at baseline. The
range of the change in PPT over 24 months at the patella was -7.35 to 7.15 kg/cm
2
; for the
wrist, the range was -6.20 to 7.28 kg/cm
2
.
Page 10 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
11
At baseline, there was synovitis in 60% of knees, effusion in 66%, and BMLs in 79% on MRI
among the whole study sample (n=1111). When evaluating the relation of these MRI lesions to
the quantitative sensory testing measurements at the patella, we found the following. Among
the whole study sample, those with synovitis at baseline had a statistically significantly lower
PPT at baseline (adjusted beta -0.42, 95% CI -0.67 to -0.18), and a statistically significant
decrease in PPT at the patella 24 months later, indicating that they had become more sensitized
(adjusted beta -0.30, 95% CI -0.52 to -0.08). However, synovitis was not associated with
temporal summation at baseline. Baseline effusion and BMLs were not associated with baseline
PPT, baseline temporal summation, or change in PPT at the patella over 24 months (Table 2).
Among those who were free of temporal summation at baseline (n=716), 22.6% exhibited
temporal summation at the patella 24 months later. In this sample, 62% had synovitis, 67% had
effusion, and 79% had BMLs on MRI at baseline. Knees with synovitis did not have an increased
risk of developing incident temporal summation at the patella (adjusted odds ratio (OR) 1.12,
95% CI 0.75-1.66, p=0.6), but those with effusion at baseline had a 54% increased risk of
incident temporal summation (adjusted OR 1.54, 95% CI 1.01-2.36, p=0.04). BML presence and
BML burden were not associated with risk of incident temporal summation (Table 3).
The relation of the MRI features to the sensory testing measurements at the wrist were as
follows. Baseline presence of synovitis was associated with lower PPT at baseline (adjusted beta
-0.19, 95% CI -0.35 to -0.03), but not with change in PPT at the wrist. In contrast, presence of
effusion was significantly associated with baseline PPT (adjusted beta -0.24, 95% CI -0.40 to -
0.07) and with decreased PPT (greater sensitivity) at the wrist over two years (adjusted beta -
Page 11 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
12
0.24, 95% CI -0.41 to -0.08, p=0.004) (Table 2). BML presence and burden were not associated
with baseline PPT or change in PPT at the wrist, consistent with the findings at the patella.
Neither synovitis, effusion, nor BMLs were associated with incident temporal summation at the
wrist (Table 3).
Finally, in exploratory analyses, persistence of synovitis over the two time-points was
associated with a decrease in PPT over the same time period at both the patella and wrist,
though not statistically significantly at the patella (patella: adjusted beta -0.24, 95% CI -0.54 to
0.06; wrist: adjusted beta -0.26 (-0.48 to -0.03)). In contrast, resolution of synovitis was not
associated with a change in PPT. Similarly, persistence of effusion was associated with a
decrease in PPT over two years (adjusted beta -0.27, 95% CI -0.46 to -0.09), but its resolution
was not associated with change in PPT. Persistence or resolution of BMLs was not associated
with change in PPT. There were no significant associations for change in any MRI feature with
change in temporal summation in the self-matched case-control study (data not shown).
Discussion
We have, for the first time, examined the relation of inflammatory lesions (synovitis and
effusion) and BMLs, which are thought to be largely reflective mechanical injury, to concurrent
and longitudinal change in pain sensitization in a large, well-characterized cohort. Congruent
with animal models of sensitization,
14-22
our findings support the potential relevance of
inflammation in the development and heightening of sensitization in knee osteoarthritis in
humans. We found that synovitis was associated with lower PPT and a decrease in PPT at the
Page 12 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
13
patella over time, indicating increased pain sensitization or sensitivity. Effusion was associated
with development of new temporal summation at the patella, and with a decrease in PPT at the
wrist, a site distant to the site of pathology; both findings suggest the involvement of central
sensitization. Thus, inflammation appears to influence the development of and perhaps
amplification of sensitization. In addition, in our exploratory analyses, we found that
persistence of synovitis and effusion were associated with a decrease in PPT over time,
primarily at the wrist, suggesting spreading sensitization over time. In contrast, resolution of
these inflammatory features on MRI did not result in a significant change in PPT, suggesting
that perhaps once sensitization or heightened sensitivity has occurred, removal of the
inflammatory stimulus may not be sufficient to alter the sensitization. This may also explain
why we found no change in temporal summation with change in these MRI features. In
contrast, BML presence and burden were not associated with either measure of sensitization.
This would imply that in human knee osteoarthritis, tissue injury per se may not influence
sensitization. Thus, BMLs appear to contribute to the pain experience through mechanisms that
are not directly related to sensitization.
The differences we found in the relations of synovitis and of effusion with our measures of
sensitization, and in the findings at the patella and at the wrist merits discussion. Inflammation
of the synovium manifests as activation of the synovial membrane, i.e., synovial thickening, or
joint effusion. In the current study, synovitis was assessed using the surrogate imaging marker
of Hoffa-synovitis (i.e., assessed as hyperintensity in the infrapatellar (Hoffa’s) fat pad on fluid-
sensitive fat-suppressed sequences). This is a sensitive but non-specific marker of synovial
Page 13 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
14
inflammation and cannot distinguish inflamed synovium from other entities such as mechanical
impingement of the fat pad or hypervascularity for other reasons.
41,33
Thus Hoffa-synovitis may
reflect local effects of inflamed synovium and/or fatty tissue that may lead to local elaboration
of adipokines that could influence local pain sensitivity.
42,43
Effusion was assessed as effusion-
synovitis, which is defined by the presence of fluid-equivalent signal in the joint cavity.
Differences in inflammatory mediators that may be relevant in effusion vs. Hoffa’s fat pad
synovitis are not well understood presently. Distension of the suprapatellar pouch may result in
effects on local pain sensitivity. Additionally, the lack of a blood-synovial barrier may potentially
implicate factors in the systemic circulation. Nonetheless, with non-contrast enhanced MRIs,
we cannot truly differentiate synovitis from effusion, and acknowledge that the signal changes
within Hoffa’s fat pad are non-specific given that the fat pad is an extrasynovial structure.
44
While BMLs are thought to primarily reflect mechanical injury and reparative attempts, we also
recognize that BMLs may have a mixed pathology, including findings consistent with localized
remodeling processes including microtrauma, fibrosis, and foci of necrosis that may have an
additional inflammatory component.
45,46
Nonetheless, unlike in rheumatoid arthritis,
inflammatory cell infiltrate does not appear to be a major feature in knee osteoarthritis.
46
Further, the results were consistently close to the null for the relation of BMLs with each of the
measures of sensitization at both the patella and the wrist.
As it is hypothesized that prolonged exposure to inflammation or tissue injury can induce
sensitization, it is possible that the time exposed to the pathologies is an important factor. It
Page 14 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
15
may be that in some (or even many) individuals, the MRI features were present for quite some
time, either persistently or in a fluctuating manner, and this could have bearing on our findings,
particularly for temporal summation. Further, our assessment of temporal summation may not
be sufficiently sensitive, as the initial pain ratings were relatively low. Indeed, temporal
summation is most frequently observed with higher initial intensity nociceptive stimuli.
Nonetheless, the differences between the two indirect measures of sensitization may also
suggest that multiple types of assessments may be needed to fully characterize different
aspects of sensitization. The combined effects of coexisting MRI lesions may provide additional
insights; however, this was beyond the scope of the current study. Additionally, there may be
pathology in other joints that we did not assess that could have contributed to pain
sensitization, thereby contributing to exposure misclassification in our study. We also
acknowledge that our findings indicate small effects; however, because factors contributing to
pain sensitization are likely multifactorial, this is not unexpected. As well, pain change over two
years is a relatively short time-frame in the broader context of knee osteoarthritis, a condition
that individuals often live with for decades.
47
We did not measure conditioned pain modulation
in our study, which may provide additional insights regarding descending inhibition of pain.
However, we were primarily interested in enhanced pain facilitation because of the
hypothesized effects of inflammation and/or injury on causing sensitization.
In conclusion, inflammation within the knee, as related to synovitis and effusion, may drive
peripheral and central sensitization in knee osteoarthritis, consistent with findings from animal
models of osteoarthritis. Since sensitization is associated with pain severity in knee
Page 15 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
16
osteoarthritis, and may potentially contribute to the transition from acute to chronic, persistent
pain in knee osteoarthritis, preventing sensitization from developing would be a potentially
effective and novel means of preventing worsening of pain in knee osteoarthritis. Early
targeting of inflammation in knee osteoarthritis may therefore be a reasonable strategy to test
for prevention of sensitization, and thereby reduction of pain severity.
Page 16 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
17
References
1. Lawrence RC, Felson DT, Helmick CG, et al. Estimates of the prevalence of arthritis and other
rheumatic conditions in the United States. Part II. Arthritis Rheum 2008;58:26-35.
2. Neogi T. The epidemiology and impact of pain in osteoarthritis. Osteoarthritis Cartilage
2013;21:1145-53.
3. Bajaj P, Bajaj P, Graven-Nielsen T, Arendt-Nielsen L. Osteoarthritis and its association with muscle
hyperalgesia: an experimental controlled study. Pain 2001;93:107-14.
4. Bradley LA, Kersh BC, DeBerry JJ, Deutsch G, Alarcon GA, McLain DA. Lessons from fibromyalgia:
abnormal pain sensitivity in knee osteoarthritis. Novartis Found Symp 2004;260:258-70; discussion
70-9.
5. Imamura M, Imamura ST, Kaziyama HH, et al. Impact of nervous system hyperalgesia on pain,
disability, and quality of life in patients with knee osteoarthritis: a controlled analysis. Arthritis
Rheum 2008;59:1424-31.
6. Lee YC, Lu B, Bathon JM, et al. Pain sensitivity and pain reactivity in osteoarthritis. Arthritis Care Res
(Hoboken) 2011;63:320-7.
7. Wessel J. The reliability and validity of pain threshold measurements in osteoarthritis of the knee.
Scand J Rheumatol 1995;24:238-42.
8. King CD, Sibille KT, Goodin BR, et al. Experimental pain sensitivity differs as a function of clinical pain
severity in symptomatic knee osteoarthritis. Osteoarthritis Cartilage 2013;21:1243-52.
9. Arendt-Nielsen L, Nie H, Laursen MB, et al. Sensitization in patients with painful knee osteoarthritis.
Pain 2010;149:573-81.
10. Arendt-Nielsen L, Egsgaard LL, Petersen KK, et al. A mechanism-based pain sensitivity index to
characterize knee osteoarthritis patients with different disease stages and pain levels. European
journal of pain 2014.
Page 17 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
18
11. Finan PH, Buenaver LF, Bounds SC, et al. Discordance between pain and radiographic severity in
knee osteoarthritis: Findings from quantitative sensory testing of central sensitization. Arthritis
Rheum 2013;65:363-72.
12. Goode AP, Shi XA, Gracely RH, Renner JB, Jordan JM. Associations between pressure-pain threshold,
symptoms, and radiographic knee and hip osteoarthritis. Arthritis Care Res (Hoboken)
2014;66:1513-9.
13. Neogi T, Frey-Law L, Scholz J, et al. Sensitivity and sensitisation in relation to pain severity in knee
osteoarthritis: trait or state? Ann Rheum Dis 2015;74:682-8.
14. Sharif Naeini R, Cahill CM, Ribeiro-da-Silva A, Menard HA, Henry JL. Remodelling of spinal
nociceptive mechanisms in an animal model of monoarthritis. Eur J Neurosci 2005;22:2005-15.
15. Schaible HG, Schmidt RF, Willis WD. Responses of spinal cord neurones to stimulation of articular
afferent fibres in the cat. J Physiol 1986;372:575-93.
16. Schepelmann K, Messlinger K, Schaible HG, Schmidt RF. Inflammatory mediators and nociception in
the joint: excitation and sensitization of slowly conducting afferent fibers of cat's knee by
prostaglandin I2. Neuroscience 1992;50:237-47.
17. Schaible HG, Schmidt RF. Effects of an experimental arthritis on the sensory properties of fine
articular afferent units. J Neurophysiol 1985;54:1109-22.
18. Grigg P, Schaible HG, Schmidt RF. Mechanical sensitivity of group III and IV afferents from posterior
articular nerve in normal and inflamed cat knee. J Neurophysiol 1986;55:635-43.
19. Schaible HG, Schmidt RF. Time course of mechanosensitivity changes in articular afferents during a
developing experimental arthritis. J Neurophysiol 1988;60:2180-95.
20. Schaible HG, Schmidt RF. Excitation and sensitization of fine articular afferents from cat's knee joint
by prostaglandin E2. J Physiol 1988;403:91-104.
Page 18 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
19
21. Schuelert N, McDougall JJ. Grading of monosodium iodoacetate-induced osteoarthritis reveals a
concentration-dependent sensitization of nociceptors in the knee joint of the rat. Neurosci Lett
2009;465:184-8.
22. Richter F, Natura G, Loser S, Schmidt K, Viisanen H, Schaible HG. Tumor necrosis factor causes
persistent sensitization of joint nociceptors to mechanical stimuli in rats. Arthritis Rheum
2010;62:3806-14.
23. Yusuf E, Kortekaas MC, Watt I, Huizinga TW, Kloppenburg M. Do knee abnormalities visualised on
MRI explain knee pain in knee osteoarthritis? A systematic review. Ann Rheum Dis 2011;70:60-7.
24. Hunter DJ, Zhang W, Conaghan PG, et al. Systematic review of the concurrent and predictive validity
of MRI biomarkers in OA. Osteoarthritis Cartilage 2011;19:557-88.
25. Hunter DJ, Guermazi A, Roemer F, Zhang Y, Neogi T. Structural correlates of pain in joints with
osteoarthritis. Osteoarthritis Cartilage 2013;21:1170-8.
26. Englund M, Felson DT, Guermazi A, et al. Risk factors for medial meniscal pathology on knee MRI in
older US adults: a multicentre prospective cohort study. Ann Rheum Dis 2011;70:1733-9.
27. Perkins BA, Olaleye D, Zinman B, Bril V. Simple screening tests for peripheral neuropathy in the
diabetes clinic. Diabetes Care 2001;24:250-6.
28. Hayashi D, Roemer FW, Katur A, et al. Imaging of synovitis in osteoarthritis: current status and
outlook. Semin Arthritis Rheum 2011;41:116-30.
29. Roemer FW, Guermazi A, Zhang Y, et al. Hoffa's Fat Pad: Evaluation on Unenhanced MR Images as a
Measure of Patellofemoral Synovitis in Osteoarthritis. AJR Am J Roentgenol 2009;192:1696-700.
30. Roemer FW, Hunter DJ, Guermazi A. Semiquantitative assessment of synovitis in osteoarthritis on
non contrast-enhanced MRI. Osteoarthritis Cartilage 2009;17:820-1; author reply 2-4.
Page 19 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
20
31. Roemer FW, Guermazi A, Felson DT, et al. Presence of MRI-detected joint effusion and synovitis
increases the risk of cartilage loss in knees without osteoarthritis at 30-month follow-up: the MOST
study. Ann Rheum Dis 2011;70:1804-9.
32. Rhodes LA, Grainger AJ, Keenan AM, Thomas C, Emery P, Conaghan PG. The validation of simple
scoring methods for evaluating compartment-specific synovitis detected by MRI in knee
osteoarthritis. Rheumatology (Oxford) 2005;44:1569-73.
33. Fernandez-Madrid F, Karvonen RL, Teitge RA, Miller PR, An T, Negendank WG. Synovial thickening
detected by MR imaging in osteoarthritis of the knee confirmed by biopsy as synovitis. Magn Reson
Imaging 1995;13:177-83.
34. Peterfy CG, Guermazi A, Zaim S, et al. Whole-Organ Magnetic Resonance Imaging Score (WORMS) of
the knee in osteoarthritis. Osteoarthritis Cartilage 2004 Mar.;12:177-90.
35. Slugg RM, Meyer RA, Campbell JN. Response of cutaneous A- and C-fiber nociceptors in the monkey
to controlled-force stimuli. J Neurophysiol 2000;83:2179-91.
36. Magerl W, Fuchs PN, Meyer RA, Treede RD. Roles of capsaicin-insensitive nociceptors in cutaneous
pain and secondary hyperalgesia. Brain 2001;124:1754-64.
37. Woolf CJ, Decosterd I. Implications of recent advances in the understanding of pain pathophysiology
for the assessment of pain in patients. Pain 1999;Suppl 6:S141-7.
38. Radloff LS. The CES-D scale: A self report depression scale for research in the general population.
Applied Psychological Measurement 1977;1:385-401.
39. Jensen MP, Keefe FJ, Lefebvre JC, Romano JM, Turner JA. One- and two-item measures of pain
beliefs and coping strategies. Pain 2003;104:453-69.
40. Leveille SG, Zhang Y, McMullen W, Kelly-Hayes M, Felson DT. Sex differences in musculoskeletal pain
in older adults. Pain 2005;116:332-8.
Page 20 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
21
41. Loeuille D, Chary-Valckenaere I, Champigneulle J, et al. Macroscopic and microscopic features of
synovial membrane inflammation in the osteoarthritic knee: correlating magnetic resonance
imaging findings with disease severity. Arthritis Rheum 2005;52:3492-501.
42. Bas S, Finckh A, Puskas GJ, et al. Adipokines correlate with pain in lower limb osteoarthritis: different
associations in hip and knee. Int Orthop 2014;38:2577-83.
43. Gandhi R, Takahashi M, Smith H, Rizek R, Mahomed NN. The synovial fluid adiponectin-leptin ratio
predicts pain with knee osteoarthritis. Clin Rheumatol 2010;29:1223-8.
44. Roemer FW, Kassim Javaid M, Guermazi A, et al. Anatomical distribution of synovitis in knee
osteoarthritis and its association with joint effusion assessed on non-enhanced and contrast-
enhanced MRI. Osteoarthritis Cartilage 2010;18:1269-74.
45. Hunter DJ, Gerstenfeld L, Bishop G, et al. Bone marrow lesions from osteoarthritis knees are
characterized by sclerotic bone that is less well mineralized. Arthritis Res Ther 2009;11:R11.
46. Zanetti M, Bruder E, Romero J, Hodler J. Bone marrow edema pattern in osteoarthritic knees:
correlation between MR imaging and histologic findings. Radiology 2000;215:835-40.
47. Losina E, Weinstein AM, Reichmann WM, et al. Lifetime risk and age at diagnosis of symptomatic
knee osteoarthritis in the US. Arthritis Care Res (Hoboken) 2013;65:703-11.
Page 21 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
22
Table 1: Participant Characteristics
Baseline Participant
Characteristics
Whole study sample
(N=1,111)
Study sample that was free
of temporal summation at
baseline (N=716)
Mean age (SD), years 66.9 (7.5) 66.3 (7.4)
% Female 62 61
Mean BMI (SD), kg/m
2
29.7 (4.8) 29.7 (4.8)
% Radiographic knee
osteoarthritis
38 37
% Frequent knee pain 21 20
Mean WOMAC pain (0-20) (SD) 2.2 (2.8) 1.8 (2.7)
Page 22 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
23
Table 2: Relation of MRI lesions to baseline PPT and change in PPT over 24 months
MRI lesion at baseline
(N=1,111)
(Prevalence)
Patella
Adjusted* beta (95% CI),
p-value
Wrist
Adjusted* beta (95% CI),
p-value
Baseline PPT Change in PPT Baseline PPT Change in PPT
Synovitis (60%)
-0.42
(-0.67 to -0.18)
p=0.0007
-0.30
(-0.52 to -0.08)
p=0.01
-0.19
(-0.35 to -0.03)
p=0.02
-0.10
(-0.26 to 0.06)
p=0.2
Effusion (66%)
0.23
(-0.03 to 0.49)
p=0.08
-0.04
(-0.28 to 0.19)
p=0.7
-0.24
(-0.40 to -0.07)
p=0.004
-0.24
(-0.41 to -0.08)
p=0.004
BML (79%)
0.14
(-0.16 to 0.44)
p=0.4
0.03
(-0.25 to 0.31)
p=0.8
0.05
(-0.14 to 0.25)
p=0.6
-0.10
(-0.29 to 0.10)
p=0.3
Sum of BMLs (“BML
burden”) (Range: 0-19)
0.03
(-0.01 to 0.08)
p=0.2
-0.01
(-0.05 to 0.04)
p=0.8
0.00
(-0.03 to 0.03)
p=0.9
0.00
(-0.03 to 0.03)
p=1.0
Page 23 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
24
*Adjusted for age, sex, BMI, race, study site, radiographic OA severity, depressive symptoms,
catastrophizing, widespread pain, and analgesic use.
Page 24 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
25
Table 3: Relation of MRI lesions to baseline temporal summation and incident temporal
summation over 24 months
MRI lesion at baseline
(Prevalence)
Patella
Adjusted* OR (95% CI),
p-value
Wrist
Adjusted* OR (95% CI),
p-value
Baseline
temporal
summation
(N=1,111)
Incident
temporal
summation
(N=716)
Baseline
temporal
summation
(N=1,111)
Incident
temporal
summation
(N=716)
Synovitis (62%)
0.83 (0.63-1.08)
p=0.2
1.12 (0.75-1.66)
p=0.6
0.84 (0.64-1.10)
p=0.2
1.22 (0.85-1.75)
p=0.2
Effusion (67%)
0.89 (0.67-1.18)
p=0.4
1.54 (1.01-2.36)
p=0.04
1.13 (0.85-1.51)
p=0.4
0.94 (0.66-1.35)
p=0.7
BML (79%)
0.95 (0.68-1.32)
p=0.8
0.92 (0.56-1.49)
p=0.7
1.04 (0.74-1.45)
p=0.8
1.05 (0.68-1.62)
p=0.8
Sum of BMLs (“BML
burden”) (Range: 0-19)
0.98 (0.93-1.03)
p=0.4
1.00 (0.92-1.07)
p=0.9
0.98 (0.93-1.04)
p=0.5
1.00 (0.94-1.07)
p=1.0
*Adjusted for age, sex, BMI, race, study site, radiographic OA severity, depressive symptoms,
catastrophizing, widespread pain, and analgesic use.
Page 25 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
26
Figure Legends
Figure 1: Examples of magnetic resonance imaging (MRI) lesion grading. Panel A depicts bone marrow
lesions (BMLs), Panels B and C depict synovitis (Hoffa-synovitis) in the intercondylar and infrapatellar
regions, respectively, and Panel D depicts effusion (effusion-synovitis) on non-contrast-enhanced MRI.
Note that BMLs are assessed on a per-subregion basis summing all BMLs in the subregion with regards
to the percent involved, while Hoffa-synovitis and effusion-synovitis are assessed on a knee basis. The
first column provides examples of grade 1 lesions, the second column provides examples of grade 2
lesions, and the final column provides examples of grade 3 lesions.
Page 26 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.
Figure 1
254x190mm (300 x 300 DPI)
Page 27 of 27
John Wiley & Sons
Arthritis & Rheumatology
This article is protected by copyright. All rights reserved.