Tendon friction rubs in systemic sclerosis:
a possible explanation—an ultrasound and
magnetic resonance imaging study
Maria S. Stoenoiu1, Fre ´de ´ric A. Houssiau1and Fre ´de ´ric E. Lecouvet2
Objective. To assess the tendon and joint involvement at wrists and ankles of patients suffering from
diffuse SSc and to identify the morphological substrate of tendon friction rubs (TFRs).
Methods. Fifteen consecutive patients suffering from diffuse SSc were included. All patients had two
musculoskeletal US (MSUS) examinations of the wrists and ankles. MRI was performed at the most
affected joints as detected by MSUS and in all sites in which TFRs were present.
Results. No clinically overt arthritis or tenosynovitis was detected in the wrists and/or ankles prior to
MSUS. Synovitis, tenosynovitis and tendon tear were identified in 8, 4 and 2 of 15 patients, respectively,
by both MSUS and MRI. At entry, 5 patients had palpable TFRs (4 bilateral and 1 unilateral) and 10 patients
did not. Tenosynovitis was more frequently found in ankles with TFRs (3/9) than in those without TFRs (3/
21), although the difference was not statistically different (P=0.3). Using MRI, deep connective tissue
infiltrates surrounding tendons were present in all sites with TFRs but in only one patient without TFRs.
Conclusion. Both MSUS and MRI are effective in detecting synovitis and tenosynovitis in diffuse SSc
patients. Tenosynovitis, synovitis and thickened retinacula are not infrequently seen in these patients. Our
data suggest that juxta-tendinous connective tissue infiltrates might be the morphological substrate of
tendon friction rubs, which may thus be a misnomer for tissue friction rubs.
Key words: systemic sclerosis, tendon friction rubs, synovitis, tenosynovitis, ultrasonography, magnetic reson-
In SSc the presence of palpable tendon friction rubs
(TFRs) is associated with disease activity and is a signifi-
cant predictor of diffuse cutaneous involvement , in-
volvement of internal organs and increased mortality .
TFRs are found in 20% of patients with established diffuse
SSc  and in 36% of patients with early diffuse SSc .
The genesis of TFRs is not clearly understood, and it is
assumed to be due to fibrin deposition inside the tendon’s
synovial sheath. However, TFRs can be palpated over the
tendons that are devoid of a synovial sheath (finger exten-
sor, quadricipital and Achilles tendons) or over the mus-
cles at a distance from the tendons. Together with TFRs,
tendon and joint involvement is common in patients with
SSc [3, 4] and is frequently disabling .
Currently, musculoskeletal involvement is assessed by
clinical examination, resulting in an underestimation of
synovitis and tenosynovitis, particularly in the presence
of increased skin thickness and flexion contractures.
Musculoskeletal US (MSUS) and MRI allow a more accur-
ate detection of synovitis and tenosynovitis than clinical
examination . The aim of the present study was to iden-
tify the morphological substrate of TFRs and to assess
the tendon and joint involvement at wrists and ankles
of patients suffering from diffuse SSc using both MSUS
1Department of Rheumatology and2Department of Radiology,
Cliniques Universitaires Saint-Luc, Po ˆle de recherche en
Rhumatologie, Institut de recherche expe ´rimentale et clinique,
Universite ´ catholique de Louvain, Louvain, Belgium.
Correspondence to: Maria S. Stoenoiu, Rheumatology Department,
Cliniques Universitaires Saint-Luc, Po ˆle de recherche en rhumatologie,
Institut de recherche expe ´rimentale et clinique, Universite ´ catholique
de Louvain, 1200 Brussels, Belgium.
Submitted 12 March 2012; revised version accepted
26 September 2012.
! The Author 2012. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: firstname.lastname@example.org
Advance Access publication 28 November 2012
SCIENCE by guest on October 21, 2015
Patients and methods
Fifteen patients suffering from diffuse SSc according to
LeRoy’s classification criteria , consecutively admitted
to the Scleroderma Outpatient Clinic, were asked and
accepted to participate in this study. The study was
approved by the institutional research ethics committee
of the Universite ´ catholique de Louvain and informed pa-
tient consent was obtained. All patients had a detailed
history and a full clinical assessment of the musculoskel-
etal system prior to the MSUS examination. Screening for
palpable TFRs was performed at the following sites:
extensor and flexor tendons of fingers, wrists, over the
elbow, shoulder, scapula, knees, ankles and toes .
Two MSUS examinations of the wrists and ankles were
performed at least 1 week apart in order to look for per-
sistent synovitis and/or tenosynovitis and both sonogra-
phers were present on the day of the second MSUS: the
first MSUS was performed using a MyLab 60 (Esaote,
Genoa, Italy) equipped with a broadband linear probe
(6?18MHz) and the second using an iu-22 system
(Philips Medical Systems, Andover, MA, USA) equipped
with a linear array broadband transducer operating with a
maximal frequency of 17.5 MHz. Abnormal findings were
confirmed on dynamic grey-scale examination using both
systems. Scans of both wrists and ankles, assessing
joints (radioulnar, radiocarpal, inter-carpal, tibiotalar), ex-
tensor retinacula and all extensor and flexor tendons at
the level of the wrists and ankles, were performed.
European League Against Rheumatism (EULAR) guide-
lines  and Outcome Measures in Rheumatology
(OMERACT) definitions  were used to evaluate the pres-
ence of synovitis/tenosynovitis. The thickness of the ex-
tensor retinacula was measured with calipers and a
thickness 590% of reported values in controls was con-
sidered abnormal [10, 11]. For MSUS assessment, inter-
observer agreement was 0.8 for synovitis and 0.9 for teno-
synovitis as calculated by Cohen’s k. Intra-observer
agreement was 0.9 for both synovitis and tenosynovitis.
MRI examinations were performed on the day of the
second MSUS and read without knowledge of the pres-
ence/absence of TFRs at wrists and ankles. In all patients,
an MRI examination was performed on the most severely
affected wrist or ankle, as identified by MSUS. In addition,
other anatomical sites in which TFRs were detected were
also imaged. Non-enhanced transverse and coronal
T1-weighted, fast spin-echo T2-weighted and short-tau
inversion recovery (STIR) and/or fat-saturated proton-
density sequences were performed  on a 1.5
Tesla scanner (Philips Medical Systems, Best, The
Netherlands) with dedicated surface coils. Standard
anterior?posterior radiographs of the wrists, hands,
ankles and feet were also obtained.
Investigation of the morphological substrate of TFRs
Fifteen consecutive patients (11 women and 4 men) with
diffuse SSc fulfilling LeRoy’s criteria were included in this
study. Mean (S.D.) age was 55 (15) years and median dur-
ation of the disease from the onset of the first non-RP
symptom was 7.5 years (range 2?18 years). None of the
patients reported current pain or swollen joints and no
clinically overt arthritis or tenosynovitis was detected in
the wrists or ankles at study entry. At entry, five patients
had palpable TFRs (4 bilateral and 1 unilateral) and 10
patients did not, without correlation between the presence
of TFRs and the presence of synovitis or increased thick-
ness of the retinacula. In contrast, tenosynovitis was most
frequently found in ankles with TFRs (3/9 or 33%) than in
those without TFRs (3/21 or 14%), although the difference
was not statistically significant (P=0.3).
Two patients with TFRs at the ankles presented with
TFRs elsewhere: one had bilateral knee TFRs at entry
and another developed bilateral TFRs on the posterior
thoracic wall soon after inclusion. These two patients
were further studied. The patient with bilateral knee
TFRs had more than four MSUS examinations performed
in the presence and in the absence of TFRs. A
fluid-containing neobursitis located at the interface of
the subcutaneous fat and the superficial aspect of the
quadricipital tendon was detected when TFRs dis-
appeared. The presence of the neobursitis was confirmed
on three different MSUS examinations (supplementary
Rheumatology Online). As soon as the fluid-containing
bursitis resolved, TFRs were palpable again over the an-
terior aspect of the knees. In another patient, the upper
girdle movement that raises and lowers the scapula over
the ribs produced friction rubs that were palpable over
both the posterior (infra-scapular regions) and lateral thor-
acic wall. MRI of these regions revealed extensive inflam-
matory infiltrates within the muscles, the fascia and at the
interface between the deep connective tissue and mus-
cles of the posterior thoracic wall (Fig. 1). The deep con-
nective tissue infiltrates resolved concomitantly with the
disappearance of the TFRs, whereas myositis underwent
a more progressive regression after the initiation of a new
These two observations led us to hypothesize that TFRs
may be due to modifications of the deep connective tissue
that impairs the smooth gliding of tendons and/or fascias.
Bearing this hypothesis in mind, we looked for the pres-
ence of deep connective tissue changes on MRI scans,
without knowledge of the TFR status of patients.
Interestingly, juxta-tendinous subcutaneous soft tissue in-
filtrates were found in all patients (5/5) in whom TFRs were
present at the ankles (Fig. 2) and in only one patient (1/10)
Joint and tendon involvement
Swollen joints were not detected by conventional radio-
graphs. Synovitis was observed in 8 of 15 patients by both
MRI and MSUS (supplementary Fig. S2A and S2D, avail-
able as supplementary data at Rheumatology Online). MRI
identified joint effusion in two additional joints (subtalar)
that were not included in the standard MSUS examination.
Tenosynovitis was identified by MSUS in 4 of 15 patients
Maria S. Stoenoiu et al.
by guest on October 21, 2015
(supplementary Fig. S2A?D, available as supplementary
data at Rheumatology Online). A positive power-Doppler
signal was present in two cases. All tenosynovitis
identified by MSUS were confirmed by MRI. Of note,
multiple-tendon compartments were always involved,
mostly bilaterally. Tendon tears were observed in two pa-
tients using both MRI and MSUS: the extensor pollicis
longus and extensor digiti communis tendons in one pa-
tient and the tibialis anterior tendon in the other (supple-
mentary Fig. S1E?G, available as supplementary data at
Rheumatology Online). In both patients, TFRs were pal-
pated over the extensor tendons despite the loss of integ-
rity of these tendons.
The main results of our study are (i) the presence of deep
connective tissue infiltrates is associated with the pres-
ence of TFRs, which might be a surrogate marker for
more extensive and deeper involvement of connective tis-
sue; (ii) the joint and tendon involvement is underesti-
mated by clinical examination; and (iii) synovitis (4 of 15)
and tenosynovitis (8 of 15) are not infrequent in patients
suffering from diffuse SSc.
First, using MRI, we observed the presence of infiltrates
in the deep connective tissue surrounding tendons of the
anterior aspect of the ankles in 6 of 15 patients. Such
changes were found in all 5 patients presenting with
TFRs but in only 1 of the remaining 10 patients without
TFRs. Interestingly, similar changes were reported by
Boutry et al.  in the deep connective tissue (along
the septae that divide fat into lobules) of the hands of
patients with SSc, but these authors did not look for a
possible correlation between these changes and the pres-
ence of TFRs. Accordingly, we suggest that TFRs may be
due to changes in the deep connective tissue, preventing
the proper gliding of tendons during active motion, rather
than to fibrinous deposits on the surface of the tendon or
tendon sheaths themselves, as previously assumed .
Cuomo et al.  found a significant association be-
tween increased thickness of the retinacula and the pres-
ence of TFRs and propose that this could be the US lesion
underlying the clinical sign. According to this hypothesis,
one cannot explain the genesis of TFRs in regions without
retinacula such as the elbow and shoulder, and around
tendons devoid of synovial sheath and retinacula, such
as the quadricipital and Achilles tendons. Moreover,
TFRs were observed at a distance from the retinacula
on the dorsal aspect of the fingers or far away from ten-
dons, above muscles such as those of the thoracic wall.
We also imaged an increased thickness of retinacula at
wrist and ankle level in 4 of 15 patients, as previously
described . However, we did not find an association
between the presence of TFRs and the thickness of the
retinacula, contrasting with the results of this previous
study . These findings are in agreement with two
recent observations of increased thickness of the pulley
system of the fingers  and of the retinacula  in pa-
tients suffering from SSc as compared with controls.
FIG. 1 MRI images of the thoracic wall in a patient with TFRs.
TFRs were palpated bilaterally over the posterior and lateral thoracic wall and disappeared after treatment. Axial
STIR sequence MRI scans were performed in the presence (A, B) and after the resolution (C, D) of TFRs, at the level of the
tip of the spine of the scapula (A, C) and infra-scapular (B, D) regions, showing extensive signal intensity changes in
the deep connective tissue, fascias and muscles of the thoracic wall. Follow-up images (C, D) show the resolution of
the changes previously observed in the deep connective tissues and partial regression of the signal intensity in the
muscles and fascias (B, D).
Tendon friction rubs in SSc
by guest on October 21, 2015
We took advantage of two interesting clinical observa-
tions of TFRs: an uncommon localization of TFRs over the
thoracic wall in one patient and a reversible pattern of knee
TFRs in the other. In the first patient, the exact location of
the neobursitis and its resolution concomitantly with the
reappearance of TFRs was helpful in identifying the anato-
mical layers involved in the genesis of TFRs: the deep con-
tendon. Another important finding is that synovial sheath or
retinacula are not required for the genesis of TFRs. In the
other patient, the uncommon localization of TFRs over the
infrascapular regions showed that TFRs can be produced
far away from the rotator cuff tendons and from tendons in
general. MRI showed extensive inflammatory changes not
connective tissue, the fascia and the muscles of the thor-
acic wall in regions over which TFRs were palpated. After
initiating immunosuppressor therapy, the inflammatory
changes observed in the deep connective tissue resolved
concomitantly with the disappearance of the TFRs,
More generally, architectural changes in the deep con-
nective tissue that modify its compliance and elasticity,
thus preventing proper gliding of its layers, might contrib-
ute to the genesis of friction rubs during active motion.
Hence we propose to replace the term tendon friction
rubs with tissue friction rubs. If confirmed, our findings
a more progressive
may be of clinical relevance, as TFRs have been consist-
ently associated with increased mortality and morbidity in
SSc patients. We hypothesize that TFRs might be a surro-
gate marker for the presence of more extensive involve-
ment of the deep connective tissue, the latter being
associated with possible involvement of internal organs.
Further prospective studies should evaluate whether the
detection of such changes in deep connective tissues sur-
rounding tendons, fascia and muscles—and possibly the
quantification of this connective tissue involvement by
(whole body) MRI or other imaging techniques such as
elastography—might allowidentification ofasubsetofpar-
ticularly at-risk patients earlier and more accurately than
TFRs, thus allowing more aggressive management.
Together with TFRs, joint and tendon involvement are
common in patients suffering from diffuse SSc. We con-
firm as previously reported that synovitis and tenosyno-
vitis are largely underestimated by clinical examination
and standard radiographs. At the wrist and hand level,
tendon and joint involvement were accurately described
in three previous studies: one using MRI  and two
others using MSUS [18, 19] as compared with radio-
graphs. Both showed the superiority of MRI and/or
MSUS versus radiographs. Chitale et al.  found that
MRI detected tenosynovitis more frequently than MSUS in
a heterogeneous subset (diffuse and limited) of eight SSc
patients with hand arthralgia. In contrast, in our study all
tenosynovitis detected by MRI were previously detected
by MSUS. To our knowledge, tendon and joint involve-
ment at the ankles in patients suffering from diffuse SSc
were not previously investigated by both MRI and MSUS.
Our pilot study has several limitations. First, patients
seen in an academic hospital such as ours might repre-
sent a subset of more severe patients. Secondly, this is a
small study and includes consecutive patients with diffuse
SSc. Accordingly, our findings may not be generalized to
all SSc patients and should be confirmed in other cohorts,
ideally in a multicentric study. Nevertheless, this series
represents a true clinical subset of SSc without overlap
syndromes. We acknowledge that standardization for
both MRI and MSUS was acquired mostly from the RA
literature, as there are no validation techniques for joint
and tendon pathology in SSc.
In conclusion, this pilot study shows for the first time
that both MSUS and MRI are able to detect joint and
tendon involvement in diffuse SSc. In addition, MRI
allows the detection of connective tissue infiltrates,
which might correspond to the anatomical substrate of
TFRs and be the hallmark of more severe connective
tissue involvement in SSc patients.
Rheumatology key messages
. Synovitis and tenosynovitis in wrists and ankles are
frequent in patients suffering from diffuse SSc.
. Both MSUS and MRI are effective in detecting
synovitis and tenosynovitis in SSc.
. In SSc patients, deep connective tissue infiltrates
might be the morphological substrate of palpable
FIG. 2 MRI images of juxta-tendinous connective tissue
infiltrates in a patient with TFRs present at the anterior
aspect of the ankles.
Sagital T1-weighted (A), T2-weighted (B) and transverse
T1-weighted (C) and fat-saturated proton-density (fsPD)
(D) MRI images showing infiltration of the subcutaneous
fat adjacent to the tibialis anterior tendon, with a low
signal on T1- and T2-weighted images (arrows in A?C) and
moderately high signal on the fsPD image (arrow in D).
Maria S. Stoenoiu et al.
by guest on October 21, 2015
The authors would like to thank all patients for their par-
ticipation, the Association des Patients Scle ´rodermiques
de Belgique. We are grateful to Dr Marie Vanthuyne and to
Mrs Genevie `ve Depresseux for their contribution to the
Belgian SSc cohort.
Funding: This work was supported in part by grants
from the Fonds pour la Recherche Scientifique en
Rheumatologie/Fonds voor Wetenschappelijke Onder-
zoek in Rheumatologie and the Fondation Saint-Luc.
Disclosure statement: The authors have declared no
conflicts of interest.
Supplementary data are available at Rheumatology
1 Steen VD, Medsger TA Jr. The palpable tendon friction
rub: an important physical examination finding in patients
with systemic sclerosis. Arthritis Rheum 1997;40:1146?51.
2 Khanna PP, Furst DE, Clements PJ, Maranian P,
Indulkar L, Khanna D. Tendon friction rubs in early diffuse
systemic sclerosis: prevalence, characteristics and longi-
tudinal changes in a randomized controlled trial.
3 Avouac J, Walker U, Tyndall A et al. Characteristics of
joint involvement and relationships with systemic inflam-
mation in systemic sclerosis: results from the EULAR
Scleroderma Trial and Research Group (EUSTAR) data-
base. J Rheumatol 2010;37:1488?501.
4 Baron M, Lee P, Keystone EC. The articular manifestations
of progressive systemic sclerosis (scleroderma). Ann
Rheum Dis 1982;41:147?52.
5 Vanthuyne M, Smith V, Arat S et al. Validation of a manual
ability questionnaire in patients with systemic sclerosis.
Arthritis Rheum 2009;61:695?703.
6 Naredo E, Bonilla G, Gamero F, Uson J, Carmona L,
Laffon A. Assessment of inflammatory activity in rheuma-
toid arthritis: a comparative study of clinical evaluation
with grey scale and power Doppler ultrasonography. Ann
Rheum Dis 2005;64:375?81.
7 LeRoy EC, Black C, Fleischmajer R et al. Scleroderma
(systemic sclerosis): classification, subsets and patho-
genesis. J Rheumatol 1988;15:202?5.
8 Backhaus M, Burmester GR, Gerber T et al. Guidelines for
musculoskeletal ultrasound in rheumatology. Working
Group for Musculoskeletal Ultrasound in the EULAR
Standing Committee on International Clinical Studies
including Therapeutic Trials. Ann Rheum Dis 2001;60:
9 Wakefield RJ, Balint PV, Szkudlarek M et al.
Musculoskeletal ultrasound including definitions for ultra-
sonographic pathology. J Rheumatol 2005;32:2485?7.
10 Robertson BL, Jamadar DA, Jacobson JA et al. Extensor
retinaculum of the wrist: sonographic characterization and
pseudotenosynovitis appearance. AJR Am J Roentgenol
11 Numkarunarunrote N, Malik A, Aguiar RO, Trudell DJ,
Resnick D. Retinacula of the foot and ankle: MRI with
anatomic correlation in cadavers. AJR Am J Roentgenol
12 Durez P, Malghem J, Nzeusseu Toukap A et al. Treatment
of early rheumatoid arthritis: a randomized magnetic
resonance imaging study comparing the effects of
methotrexate alone, methotrexate in combination with
infliximab, and methotrexate in combination with
intravenous pulse methylprednisolone. Arthritis Rheum
13 Boutry N, Hachulla E, Zanetti-Musielak C, Morel M,
Demondion X, Cotten A. Imaging features of musculo-
skeletal involvement in systemic sclerosis. Eur Radiol
14 Shulman L, Kurban A, Harvey A. Tendon friction rubs
in progressive systemic sclerosis (scleroderma). Trans
Assoc Am Phys 1961;74:378?88.
15 Cuomo G, Zappia M, Iudici M, Abignano G, Rotondo A,
Valentini G. The origin of tendon friction rubs in patients
with systemic sclerosis: a sonographic explanation.
Arthritis Rheum 2012;64:1291?3.
16 Tagliafico A, Panico N, Serafini G, Ghio M, Martinoli C.
The thickness of the A1 pulleys reflects the disability of
hand mobility in scleroderma. A pilot study using
high-frequency ultrasound. Eur J Radiol 2011;77:254?7.
17 Low AH, Lax M, Johnson SR, Lee P. Magnetic resonance
imaging of the hand in systemic sclerosis. J Rheumatol
18 Cuomo G, Zappia M, Abignano G, Iudici M, Rotondo A,
Valentini G. Ultrasonographic features of the hand and
wrist in systemic sclerosis. Rheumatology 2009;48:
19 Elhai M, Guerini H, Bazeli R et al. Ultrasonographic hand
features in systemic sclerosis and correlates with clinical,
biological and radiographic findings. Arthritis Care Res
2012, doi: 10.1002/acr.21668.
20 Chitale S, Ciapetti A, Hodgson R et al. Magnetic reson-
ance imaging and musculoskeletal ultrasonography detect
and characterize covert inflammatory arthropathy in sys-
temic sclerosis patients with arthralgia. Rheumatology
Tendon friction rubs in SSc
by guest on October 21, 2015