Article Text

Extended report
Prevalence of ultrasound synovial inflammatory findings in healthy subjects
  1. Ilaria Padovano1,
  2. Félicie Costantino1,2,
  3. Maxime Breban1,2,
  4. Maria Antonietta D'Agostino1,2
  1. 1Rheumatology Department, APHP, Hôpital Ambroise Paré, Boulogne-Billancourt, France
  2. 2INSERM U1173, Laboratoire d'Excellence INFLAMEX, UFR Simone Veil, Versailles-Saint-Quentin University, Saint-Quentin en Yvelines, France
  1. Correspondence to Professor Maria Antonietta D'Agostino, Rheumatology Department, Hôpital Ambroise Paré, 9 avenue Charles de Gaulle, Boulogne 92100, France; maria-antonietta.dagostino{at}


Objective To evaluate the prevalence of joint inflammatory abnormalities and erosions detected by grey-scale and Doppler ultrasound (US) in the small joints of hands and feet in healthy subjects.

Methods US of the dorsal surface of 32 joints (10 metacarpophalangeal, 10 proximal interphalangeal, 10 metatarsophalangeal (MTP) and 2 wrists) was performed in 207 healthy subjects without joint symptom. Synovial effusion (SE), synovial hypertrophy (SH) and power Doppler (PD) signal were scored using a semiquantitative grading scale (0–3) and erosion binary.

Results One-hundred and eighty-two subjects had at least one US abnormality: 52% of the subjects had SE alone, 13% SH alone (5% with and 8% without PD) and 35% both SH and SE. US findings were detected in 9% of the total joints examined, mostly in the feet, and in particular in the MTP1 (33% of the positive joints). SE was the most frequently detected finding (68% of the positive joints), followed by SH (31%). Severity was mild (grade 1 in average) whatever the finding recorded (SH, SE or PD). Four erosions were detected (MTP1).

Conclusions This study describes for the first time, in a large cohort of healthy subjects, the prevalence and location of US signs of joint inflammation and of structural damage in small joints of hands and feet. US abnormalities were quite common, and mostly located in the feet. Further studies are needed to define which US components may allow to discriminate between pathological and physiological findings in the joints commonly affected by inflammatory arthritis conditions.

  • Ultrasonography
  • Synovitis
  • Inflammation
View Full Text

Statistics from

Over the last 15 years, musculoskeletal ultrasound (US) has become a widely used imaging technique for the evaluation of joint involvement in inflammatory rheumatic diseases, especially for the assessment of synovitis in patients with rheumatoid arthritis (RA).1 Several studies have shown that US is more sensitive than clinical examination for detecting joint inflammation in any phase of the RA management: diagnosis, follow-up and even when clinical remission is achieved (ie, ‘subclinical inflammation’).2–4 In addition, US has shown to be more sensitive than conventional radiography for detecting articular damage (ie, erosions),5 ,6 with a comparable accuracy for detecting signs of inflammation as MRI, while less sensitive for detecting erosions.7–10

Despite these data, the lack of standardisation and the long training are still some of the major criticisms to the technique, which hamper a larger use of this tool in the daily management of RA and its implementation as an outcome measure in therapeutic clinical trials. To improve the acquisition and interpretation of US images, several initiatives have been undertaken. In 2001, the EUropean League Against Rheumatism (EULAR) Imaging working group has established a consensual acquisition protocol for individual joints.11 In 2005, the Outcome Measure in Rheumatology (OMERACT) US group has proposed the first consensus-based definitions of a US-detected synovitis and of individual components of joint inflammation: synovial hypertrophy (SH) and synovial effusion (SE) in B-mode or grey scale (GS) and synovial vascularisation assessed by power Doppler (PD).12 At the same time, the group performed several consecutive studies in order to improve the reliability of the technique and to develop a standardised scoring system for appropriately quantifying RA synovitis.13–19 They developed new definitions of each single component (ie, SH, SE and PD) and proposed a novel composite (GS and PD) 0–3 semiquantitative global score of synovitis: the PDUS (power Doppler US) GLOSS (GLobal OMERACT Scoring System), which considers only the hypoechoic SH and the intrasynovial Doppler signal, independently of the presence of SE (anechoic finding), for the definition of synovitis, using the highest score between GS and PD for defining the final grade.18

However, a number of questions still remain unanswered: the concept of minimal activity of a US-detected synovitis at the joint level, as well as the lack of information on what constitutes a ‘normal’ level of GS and PD findings in healthy subjects, without joint symptoms. The last point is also related to the quality of the US machines and the continuous improvement in the technology making them able to detect even minimal changes.20 ,21 Consequently, the clinical relevance of what is considered a ‘positive’ GS and PD finding needs to be more precisely clarified in order to improve the US specificity and prevent false-positive results.

The aim of this study was therefore to determine the prevalence of a US-detected synovitis and of each single component (ie, SE and SH in GS and intrasynovial PD signal), as well as the presence of intra-articular bone erosions in healthy subjects, in order to discriminate between physiologic and pathologic US findings at hands and feet joint level.

Subjects and methods


Between October 2013 and March 2014, healthy volunteers were consecutive recruited among (1) healthcare medical, paramedical and administrative staff of the Ambroise Paré Hospital in Boulogne-Billancourt (France); (2) students in medicine or nursing frequenting the same hospital; (3) healthy relatives visiting or accompanying patients and (4) volunteers enrolled through a web announcement of that study. Written informed consent was obtained from all participants. Inclusion criteria were: age from 18 to 90 years and free consensus to participate in the study. Exclusion criteria were: (1) personal history of RA and of any rheumatic inflammatory or degenerative joint diseases including osteoarthritis of the hands and feet, (2) history of joint trauma or of enteric and/or genitourinary infection in the last month, (3) joint pain (whatever and wherever) during the last month and (4) clinically detectable joint inflammation at physical examination. In addition to the age, the following data were collected: gender, ethnicity, professional activity (notably hand work), sports and leisure activities, with a focus on sports and hobbies that could lead to mechanical stress of the hands and/or feet.

All subjects completed a visual analogue scale (VAS) evaluation from 0 to 100 as for their level of joint pain over the last month, and were asked to declare any intake of analgesic and/or non-steroidal anti-inflammatory drugs for joint pain during the last month.

Clinical examination

All subjects underwent a physical examination of the same joints assessed by US by an independent rheumatologist in order to detect the presence of any tender and/or swollen joint.

Only those subjects who did not present any tender and/or swollen joint and who were asymptomatic (VAS less than 10/100) for joint pain in the month prior to the enrolment were included into the study.

US assessment

US scanning was performed using a MyLab70 XVG (Esaote, Genoa Italy), equipped with a 6–18 MHz broadband multifrequency linear array transducer; GS settings were standardised to a frequency of 18 MHz and PD was set with a minimal frequency of 11 MHz. The pulse repetition frequency was adapted to small vessels with slow flow (around 750 Hz). The Doppler gain was adjusted to a level just below the disappearance of background noise (between 40% and 60%); the window of Doppler was placed on the whole joint reaching the surface of the skin to avoid artefacts related to the reverberation of superficial vessels. A generous quantity of gel was used.

Multiplanar bilateral US examination of the dorsal surface, in both longitudinal and transverse scans, of wrist, metacarpophalangeal (MCP) from 1 to 5, proximal interphalangeal (PIP) from 1 to 5 and metatarsophalangeal (MTP) joints from 1 to 5 was performed, for a total of 32 examined joints per subject. Wrist was examined at three compartments: radiocarpal, ulnocarpal and intercarpal, while the second and fifth MCP joints, as well as the first and fifth MTP joints were also assessed using the lateral scan. The choice to examine only the dorsal aspect of the joints and not to include the palmar aspect was mainly based on the fact that the majority of the published studies used such approach for evaluating joint inflammation in RA.

We searched for the presence of synovitis and of each elementary component (SH, SE and PD) as well as for the presence of US erosions. Synovitis was described according to the OMERACT definition12 as a hypoechoic intra-articular tissue (ie, GS SH) not displaceable and poorly compressible and which may exhibit intrasynovial PD signal. The presence of SE was defined as an anechoic intra-articular area, displaceable and compressible, and clearly different from SH (anechoic–hypoechoic) and which does not exhibit Doppler signal. Each single component (ie, SH, SE and PD) was scored separately using a validated semiquantitative (0–3) grading scale (figure 1).19 US erosion was defined as an intra-articular discontinuity of the bone surface visible in two perpendicular planes, according to the OMERACT definition.12 The US examination was performed by a sonographer with 2 years of experience in musculoskeletal US (IP), trained by a senior with 20 years of US experience (MADA). The average duration of the US examination was 30 min. Doubtful US images were revaluated by the senior.

Figure 1

Overview of the semiquantitative grading for grey-scale synovial effusion (SE), grey-scale hypoechoic synovial hypertrophy (SH) and power Doppler (PD) signal within the synovium. Examples are from the dorsal surface of the first metatarsophalangeal joint (MTP1).

Statistical analysis

Statistical analysis was performed using GraphPad Prism V.5.00 version for Windows (GraphPad Software, California, USA). The frequency of dichotomous variables (gender, work and sport/hobby activities) in the two groups (positive and negative by US) was analysed using 2×2 contingency tables using the Fisher's test. Since the age distribution in both groups did not satisfy the condition of normality, this variable was compared using the non-parametric Mann–Whitney test.


Characteristics of the population

We recruited 219 subjects, and 207 of them were finally included into the study (173 Caucasian, 30 Afro-Caribbean and 4 Asian). Ten subjects were excluded because of previous traumatisms and/or presence of joint pain (VAS >10) in more than one of the joints examined. Two subjects were excluded for ongoing treatment with potential impact on joint inflammation: (1) antitumour necrosis factor α inhibitor in relation with an active Crohn's disease and (2) azathioprine for an autoimmune hepatitis. The mean age of the included population was 35.5±12.8 years (range from 18 to 74 years), and 61 (29%) subjects were men. Seventeen subjects had a family history of psoriasis, 8 of RA, 1 of both RA and Sjögren's syndrome and 1 of systemic sclerosis.

US findings

US findings at population level

One hundred and eighty-two out of the 207 included subjects (88%) presented at least one joint with US abnormal finding: 95 of them (52%) presented SE alone, while the 87 remaining subjects (48%) presented synovitis (ie, SH with or without PD signal). In 64 out of those 87 subjects (74%), synovitis was associated with SE and in 8 cases (9%) SH was associated with PD (figure 2).

Figure 2

Prevalence of US findings at subject (A) and joint (B) levels. SH, grey-scale hypoechoic synovial hypertrophy; SE, grey-scale synovial effusion; PD, power Doppler signal within synovium; US, ultrasound.

Distribution of gender, work and sport/hobby activities was not significantly different between the group of individuals presenting US positive findings and the group without any US finding. Age was higher in the group with US findings (mean±SD: 40±14 years; median: 39 years; range: 18–65 years) than in the group without any US finding (mean±SD: 34±12 years; median: 31 years; range: 18–74 years) with a significant p value (p=0.0008), and SE was more frequent in the youngest group (18–29 years) while SH was predominant in the oldest group (50–59 years) (see online supplementary tables S1 and S2).

US findings at joint level

Three wrists of three subjects were not scanned because of past history of trauma (more than 15 years earlier in each case) resulting in 6621 joints examined by US. When considering the total number of joints examined, 9% (586/6621) presented a US positive finding: 69% (403/586) presented SE alone, 31% (183/586) synovitis (SH with or without intrasynovial PD) and 1% (6/586) both SE and SH (3 of which were also PD positive). US erosions were detected in four joints (all in the MTP1, medially), and their presence was always associated with the presence of SE (figure 2).

Preferential location of US-detected abnormalities

Finger and wrist joints showed a much lower prevalence of US abnormalities as compared with MTP joints (544 out of the 586 involved joints). MTP1 was the most frequently involved (36%), followed by the 2°, 3°,4° and 5° MTPs (30.5%, 22%, 11% and 0.5%, respectively). In the upper limbs, radiocarpal joint was involved in 6% of the examined wrists, followed by MCP3 (2%). No abnormality was detected in the MCP5 or in any PIP joints. The findings were predominant at the dominant hand and at the left side of the feet. Detailed results are reported in tables 1 and 2.

Table 1

Distribution of synovial effusion, synovial hypertrophy and intrasynovial power Doppler signal at joint level

Table 2

Preferential location of abnormal US finding and related grading

Whatever the US component (SE, SH or PD), the most frequently detected severity score was grade 1, which was recorded in 94% of the MTP1 SE-positive joints, in 65% of those with SH positive and in 74% of those with positive PD. A similar trend was observed for MTPs 2, 3 and 4. MTP1 and MTP2 were the only joints to display grade 3 of SH (tables 1 and 2).


To our best knowledge, this is the first study evaluating systematically in a large cohort of healthy subjects the prevalence of US abnormalities corresponding to the presence of inflammation (ie, SH with or without intrasynovial PD signal and SE) and of structural damage (ie, erosions) in the joints typically involved in the course of RA. Several important results can be underlined. First, the frequency of US abnormalities was relatively low at the joint level (9% of the examined joints) whereas it was high at the subject level (88% of the studied population presented at least one US abnormality). Moreover, the most frequently detected US abnormality, both in terms of number of subjects and of joints involved (52% and 69%, respectively) was SE. A clear sign of synovitis (SH with or without PD) was found in a relatively high proportion of the involved subjects in at least one joint (48%). However, the severity of the detected findings was most generally mild (grade 1). Finally, the preferential location of those abnormalities was the MTPs, especially the first.

Some studies have already shown that US findings suggestive of synovial inflammation may occur in healthy subjects,9 ,22–25 all corresponding to a lower number of examined subjects. Our observation that the most frequently recorded US abnormality in healthy subjects was SE may indirectly support the choice made by the OMERACT US group to exclude this component in the definition of a US synovitis.18 Conversely, SH was less frequently found, and even less when associated with intrasynovial PD signal, suggesting that those abnormalities were more likely to be specific for synovial inflammation.

The observed distribution of the US findings in healthy subjects may contribute to establish a selection of the joints to be examined for diagnostic purpose. Indeed, the high frequency of US abnormalities of the MTP1, 2 and 3 in healthy subjects, also identified by Ellegaard et al25 and Kitchen and Kane,24 might lead to exclude those joints from a global joint count, both for early diagnosis purpose and for disease activity and treatment response monitoring, or on the other hand, to define a threshold of normality.

The preferential location of abnormal findings in MTP joints can possibly be explained by biomechanical factors (joint loading), leading to microtrauma and subclinical stress of those joints, as well as by the possible combined presence of local asymptomatic osteoarthritis changes especially with increasing age. Therefore, the observed joint inflammatory findings in healthy subjects are more likely to be explained by biomechanical factors rather than by an inflammation of the synovial membrane of immunological origin, such as in RA. The low percentage of abnormalities found in healthy subjects in the other examined joints (PIP, MCP, wrist and MTP5), may probably result in a higher specificity of joint inflammation when present.

Another ‘practical’ point to emphasise for diagnostic purpose was the mild severity of US abnormalities detected in healthy subjects. Millot et al,22 and Witt et al,23 found that grade 1 of GS was the most represented US findings leading them to suggest that this grade should not be considered as pathological. Our results support their observations. Despite the high prevalence of subjects with US abnormalities, a possible limitation is the fact that the examiner was aware of including only healthy individuals without joint pain or disease.

The low frequency of US abnormalities at joint level in our study and their mild severity could be explained by the relatively young age of the included subjects (mean of 35.5 years) as compared with usual age of patients with RA. Age should be considered to interpret US joint examination since we found a significant increase of US abnormalities with the increase of the age.

In conclusion, this study, conducted in a large sample of healthy population, confirms that US abnormalities suggestive of inflammation (ie, SH, SE and intrasynovial PD) can be detected in healthy individuals. Such abnormalities were more frequent in the MTPs and increased with age. Our results support the decision to exclude SE findings as for the diagnostic assessment of RA and the monitoring of disease activity because of its relatively high prevalence observed in healthy subjects. For the same reason, the detection of abnormalities at MTP level (especially MTP1), as well as the detection of abnormalities of low grade of severity (grade 1) should be considered with caution when used for diagnostic purpose. Further studies are needed to define a US threshold of abnormalities and to test its relevance in suspected arthritis.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Tore K Kvien

  • Contributors All the individuals listed as authors fulfil the uniform authorship credit requirements for manuscripts submitted to medical journals, that is, that they all contributed to the manuscript based on substantial contributions to conception and design, acquisition of data or analysis and interpretation of data; drafting the article or revising it critically for important intellectual content and final approval of the version to be published.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.