Ultrasonography in Early Rheumatoid Arthritis of Hand and Wrist Joints: Comparison with Magnetic Resonance Imaging

Abstract

Background

The aim of the study was to evaluate the use of ultrasonography (USG) including power Doppler in detecting hand and wrist joint changes in early rheumatoid arthritis (RA) and to compare USG findings with magnetic resonance imaging (MRI).

Materials and Methods

Thirty-four patients diagnosed as RA by 2010 ACR/EULAR criteria; with the onset of symptoms within last one year, were included in the study after institute ethical clearance and informed consent to undergo USG and contrast-enhanced MRI of the dominant affected hand. Second to fifth metacarpophalangeal (MCP) joints, second to fifth proximal interphalangeal (PIP) joints and wrist joints (total nine joints) were evaluated for synovitis, erosions along with tenosynovitis. USG and MRI features were compared; agreement on the two imaging modalities as well as sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy of ultrasonography compared to MRI (gold standard) was calculated.

Results

One hundred thirty-six MCP, 136 PIP and 34 wrist joints (total 306 joints) and 136 flexor tendons were evaluated. The sensitivity, specificity, PPV, NPV and diagnostic accuracy of USG for diagnosing synovitis was 78.6%, 91.1%, 86.1%, 85.8%, 86.3%; for erosions 67.2%, 97.5%, 84.8%, 90.5%, 91.5%; for tenosynovitis 86.5%, 100%, 100%, 92.3% and 94.8% respectively. The overall agreement between USG and MRI for detection of synovitis was achieved in 83% joints and for erosions in 89.5% joints.

Conclusion

In early RA, USG was nearly as effective in diagnosing features of joint and tendon sheath involvement, with relatively better performance of USG for tenosynovitis. The performance of USG in diagnosing erosions was limited likely due to difficult access of three-dimensional joint structure.

Background

Rheumatoid arthritis (RA) is an autoimmune systemic disorder characterised by inflammatory synovitis of peripheral joints. The disease affects females more commonly than males with prevalence ranging from 0.1 to 5% [1, 2]. Timely initiation of appropriate therapy and treatment monitoring is of paramount importance as research has shown that treatment during the window period may prevent further joint damage and possibly reverse the same [3, 4].

Ultrasonography (USG) has been increasingly used to assess the articular and extra-articular musculoskeletal manifestations of the disease. It has been used not only to diagnose and quantify the joint disease but also in follow-up and evaluation of treatment response in these patients. Magnetic resonance imaging (MRI) is considered the gold standard in the evaluation of RA, especially when the histopathological correlation is uncommon [5]. However, USG is a more widely available imaging tool that can be used to evaluate early articular and extra-articular manifestations and has demonstrated a role in joint abnormality assessment [6–9].

Therefore, this study was performed to evaluate the use of USG in detecting hand and wrist joint changes in early RA and to compare the USG findings with that of MRI.

Materials and Methods

Study Design

Prospective crossectional study.

Patient Population

Thirty-four consecutive patients of RA, presenting to the rheumatology outpatient department from April 2014 to April 2016, aged more than 12 years, with a score of more than 6/10 (ACR/EULAR 2010 criteria) [10], with the onset of symptoms within the past one year were included in the study to undergo USG and contrast-enhanced MRI of the dominant affected (with more number of joints involved clinically) hand and wrist. Clinical involvement was defined as the presence of either swelling with tenderness or swelling alone over a joint. Second to fifth metacarpophalangeal (MCP) joints, second to fifth proximal interphalangeal (PIP) joints and wrist joints (total nine joints) were evaluated for synovitis and erosions along with tenosynovitis of the second to fifth flexor tendons.

Image Acquisition and Analysis

All ultrasonography examinations, were performed by one radiologist (S.V. with 15 years of experience or S.M. with 2 years of experience), blinded to the MRI findings, who was trained in musculoskeletal sonography. The examinations were done on Aixplorer supersonic ultrasound system using a linear (7.5–15 MHz) probe, jelly mound and minimal pressure. The patients were in sitting position with forearms supported in front of them to maximize comfort. All the patients were evaluated for by ultrasound for synovitis, erosions and tenosynovitis using the following definitions.

Synovitis Uncompressible hypoechoic intracapsular area extending to fill the angle between the periarticular bones which may bulge over the line linking the top of the articular bones. Power Doppler findings were also recorded, however, these were not separately analysed.

Bone Erosion Break in the cortical outline of bone in the area adjacent to the joint, visualized in two planes measuring at least 2 mm.

Tenosynovitis Hypoechoic or anechoic area in the tenosynovium surrounding the tendon.

MRI for all patients was performed on a 1.5 T Philips Achieva system using a dedicated, small 8 channel surface coil. The hand was placed in the superman position or in supine hand by side if the former was found uncomfortable. The MRI protocol included STIR TSE (coronal), T1W TSE (coronal), DWI (coronal), multi fast field echo (mFFE 0.5(coronal), T2W Spectral Attenuated Inversion Recovery (SPAIR-axial) and T1W TSE (axial). These were followed by intravenous gadolinium (dose 0.1 mmol/kg body weight). T1W TSE Post-contrast (coronal and axial) images were then obtained. The hand was evaluated for the presence of synovial thickening, synovitis, erosion and tenosynovitis on both the modalities. Standardized MRI definitions, summarized below were used as provided by the OMERACT RAMRIS [10].

Synovitis An area of synovium that shows above normal post-gadolinium enhancement, the thickness of which is greater than the width of the normal synovium (more than 0.5 mm).

Bone erosion A sharply marginated periarticular bony lesion, visible in two planes with a cortical break seen in at least one plane, appearing hypointense on T1W images, hyperintense on T2W images and shows enhancement on post-gadolinium images.

Tenosynovitis An area of T2 hyperintense signal in the tendon sheath and/or increased thickness or contrast enhancement of the tendon sheath synovium.

Statistical Analysis

The data was recorded in a predesigned proforma and entered on excel spreadsheet in Microsoft excel [11]. The data were analysed using STATA 12.1 software [12]. Descriptive statistics were expressed as mean and standard deviation for data sets without any extreme values. Diagnostic test [13] was used to calculate the agreement as well as the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy of USG for evaluation of erosions as compared to MRI, which was taken as the gold standard. Similarly, the agreement, as well as the sensitivity, specificity, PPV, NPV and diagnostic accuracy of clinical examination and ultrasonography, was calculated as compared to MRI using the diagnostic test.

Results

Out of a total of 34 patients, 28 were female and 6 were males, with a mean age of 35.7 + 10.1 years (range 19–65 years). 136 MCP, 136 PIP and 34 wrist joints (total 306 joints, 9 per hand) and 136 flexor tendons (4 per hand) were evaluated.

Synovitis

Synovitis was found in 104 joints on clinical examination, 133 joints on ultrasonography and 145 joints on MRI. The most commonly involved joints with synovitis were the wrist, MCP 2, MCP 3 and MCP 5 in decreasing order. The sensitivity, specificity, PPV, NPV and diagnostic accuracy of USG for diagnosing synovitis was 78.6%, 91.1%, 86.1%, 85.8%, 86.3%. Agreement between ultrasonography and MRI was achieved in 83% joints, whereas it was achieved for 81.4% joints between clinical examination and MRI [Table 1].

Table 1.

Distribution of synovitis among hand and wrist joints on clinical examination (CE), ultrasonography (USG) and MRI with performance parameters of CE and USG

Joints	No. of joints			Agreement (%)		Sensitivity (%)		Specificity (%)		PPV (%)		NPV (%)		Diagnostic accuracy
	USG	MRI	CE	USG-MRI	CE-MRI	USG	CE	USG	CE	USG	CE	USG	CE	USG	CE
MCP2	17	19	13	88.2	76.5	84.2	63	93.3	93	94.1	92	82.3	67	88.2	76
MCP3	19	18	16	85.3	82.3	88.9	77.8	81.2	87.5	84.2	87.5	86.7	77.8	85.3	82.3
MCP4	10	13	8	79.4	79.4	61.5	53.8	90.5	95.2	80	87.5	79.2	76.9	79.4	79.4
MCP5	13	19	10	73.5	73.5	63	52.6	93	100	92	100	66.7	62.5	76.5	73.5
PIP2	9	10	9	91.2	91.2	80	80	95.8	95.8	88.9	88.9	92	92	91.2	91.2
PIP3	12	13	9	70.6	70.6	77	46.1	90.5	85.7	83.3	66.7	86.4	72	85.3	70.6
PIP4	10	11	4	79.4	79.4	72.7	36.4	91.3	100	80	100	87.5	76.7	85.3	85.3
PIP5	11	10	5	85.3	85.3	80	50	87.5	100	72.7	100	91.3	82.8	85.3	85.3
Wrist	32	32	30	94.1	94.1	100	93.7	100	100	100	100	100	50	100	94.1
Total	133	145	104	83	81.4	78.6	61.5	91.1	93.4	86.1	91.4	85.8	82.4	86.3	82

Erosions

USG detected erosions in 46 joints, whereas MRI detected in 60 joints. Out of the individual joints evaluated, the highest number of erosions were detected in the wrist (25 joints), followed by MCP 3 (9 joints), MCP 2 (8 joints) and MCP 5 (7 joints) on MRI. The sensitivity, specificity, PPV, NPV and diagnostic accuracy of ultrasonography for diagnosing erosions was 67.2%, 97.5%, 84.8%, 90.5%, 91.5% respectively. The overall agreement between ultrasonography and MRI for the detection of erosions was achieved in 89.5% joints [Table 2].

Table 2.

Distribution of erosions among hand and wrist joints on ultrasonography (USG) and MRI with USG performance parameters

Joints	No. of joints on USG	No. of joints on MRI	USG-MRI agreement (%)	Sensitivity of USG (%)	Specificity of USG (%)	PPV USG (%)	NPV USG (%)	Diagnostic accuracy USG (%)
MCP2	4	8	82.3	57.5	96.1	75	83.3	82.3
MCP3	7	9	82.3	55.6	92	71.4	85.2	82.3
MCP4	4	4	88.2	50	93	50	93.3	88.2
MCP5	4	7	91.2	57.1	100	100	90	91.2
PIP2	1	1	100	100	100	100	100	100
PIP3	3	2	97.1	100	96.9	66.7	100	97.1
PIP4	1	2	97.1	50	100	100	97	97.1
PIP5	1	2	97.1	50	100	100	97	97.1
Wrist	21	25	88.2	84	100	100	69.2	88.2
Total	46	60	89.5	67.2	97.5	84.8	90.5	91.5

Tenosynovitis

Out of the 136 flexor tendon sheaths examined, tenosynovitis was detected in 45 tendons on ultrasonography and 52 tendons on MRI. The sensitivity, specificity, PPV, NPV and diagnostic accuracy of ultrasonography for diagnosing tenosynovitis was 86.5%, 100%, 100%, 92.3% and 94.8% respectively.

Discussion

Imaging in RA is a challenging field with limited evidence regarding clear cut indications in diagnosis and follow up. For early detection, various imaging modalities may help before the onset of irreversible changes, which can benefit the patients by preventing significant long term morbidity and disability. Our study evaluated 34 patients with early RA using clinical examination, ultrasonography and MRI. 82% of these were females with a F:M ratio of 4.5:1 which is similar to other Indian data [14]. The mean age was 35.7 years with 50% of the patients in the age group of 31–40 years. Most of our patients were young (< 60 years), and small joints of the hand are more affected in these patients [15, 16].

Synovitis is one of the earliest manifestations of the disease. The sensitivity of USG in detecting synovitis was more than that of clinical examination (78.6% vs 60%) with no loss of specificity (Fig. 1). Previous studies have also shown similar results and have also suggested that the low sensitivity of clinical examination may account for the deterioration of RA patients despite clinically adequate control of the disease [5, 17]. The present study shows the sensitivity of USG in diagnosing signs of synovial inflammation to be 78.6% compared to MRI (gold standard). This is lower than the results of a previous study where the sensitivity of ultrasound was found to be almost equal to MRI [18]. This can be due to a higher resolution of 1.5 T MRI with a multichannel extremity coil used in our study, compared to a 0.5 T MR scanner used in the study by Schmidt WA et al. We also found six joints where synovitis was seen on greyscale USG but not on MRI, which could be due to fibrotic pannus, better visualised on USG. However, all our patients were of early RA, wherein the presence of fibrotic pannus is relatively uncommon. A similar discrepancy was also in a previous study [19].

Fig. 1.

Correlation of joint involvement on USG and MRI. A young female patient with pain in the metacarpophalangeal (MCP) joints and wrist right on the right side. Clinical examination (a) and AP radiograph (b) of the right hand revealed no abnormality. Longitudinal USG (c) through MCP4 joint shows hypoechoic synovial proliferation (arrow), with similar findings on post-contrast T1W fat-suppressed axial and coronal MRI images (d) as enhancing synovium

Erosions are an important manifestation of the joint disease in RA and are included in the diagnostic criteria for RA. Their presence predicts a more severe clinical course and increased morbidity [20]. Literature has already shown a higher sensitivity of USG in diagnosing erosions as compared to clinical examination and radiography, in part due to inherent advantages of the multidimensional capability of USG compared to a two-dimensional projection of a radiograph [5] (Fig. 2). In the present study, USG had a sensitivity of 67% in detecting erosions with a high specificity of more than 97% (Fig. 3). The increased sensitivity from previously reported of around 60% can be attributed to improvement in USG technology [5]. The sensitivity was highest for PIP joints followed by MCP 2 joints, likely due to better accessibility as compared to MCP joints [19, 21]. USG also had two false positives, where erosions in PIP joints were not seen on MRI. The slice thickness of MRI sequences was 3 mm, which may be suboptimal for small erosions at PIP joints. The highest number of erosions were detected in wrist followed by MCP3 and MCP 2 joints as these are the earliest joints to be involved in RA. This is especially relevant with the abbreviated 12 joint and 7 joint ultrasonography scores which reduce the duration of examination without compromising the sensitivity and specificity [22].

Fig. 2.

Erosions missed on radiography with correlation on USG and MRI. A young male patient with wrist pain and morning stiffness. Wrist AP radiograph (a) showed mildly reduced radiocarpal joint space and periarticular osteopenia. Sonography at the distal radius (b) showed an erosion (arrow) with adjacent synovitis. The erosion was confirmed on two planes. Corresponding axial T1W MRI (c) and post-contrast fat-saturated T1W MRI (d) image showed a well- defined cortical defect on the volar surface of the distal end of the radius (white arrow) that appears hypointense on T1W image with contrast enhancement

Fig. 3.

Limitation of USG in the diagnosis of erosions. A young female patient with positive rheumatoid factor, presented with localized pain in MCP3 joint of the left hand. Sonography (a) showed a normal MCP3 joint with no erosions or synovitis. Corresponding T2W (b) and T1W (c) MRI images showed the presence of a small erosion (arrow) on the radial aspect of MCP3 with associated soft tissue thickening. The location of erosion at the lateral or medial metacarpal heads may make it inaccessible to USG probe

Tenosynovitis of the flexor tendon may be seen in more than 50% of patients of known RA and has also shown to be an early predictor of the disease [23, 24] (Figs. 4, 5). We evaluated 136 flexor tendons for tenosynovitis with USG and MRI, and the sensitivity and specificity of ultrasonography were found to be 86.5% and 100% respectively. In two of our subjects, tenosynovitis was the only imaging finding. These subjects had arthritis of large joints with no affliction of the small joints of the hand. Positive serology in the form of elevated RA factor and anti-CCP was helpful in making a diagnosis. It is shown by previous researchers that tenosynovitis of the common flexor tendons can be the earliest manifestation of RA and that tenosynovitis of the flexor tendons of the second finger is significantly associated with progression to RA [23, 24].

Fig. 4.

Tenosynovitis as the earliest manifestation of RA. A middle-aged female with swelling along the base of right thumb and wrist. RF factor and ESR were elevated. Radiograph of the right hand (a) showed mild soft tissue thickening at medial aspect of the wrist. Transverse sonography (b) of the first extensor compartment tendons showed hypoechoic soft tissue along the synovial sheath of abductor pollicis longus and extensor pollicis brevis. MEDIC coronal (c) and post-contrast axial (d) MRI images of wrist showed marked hyperintense soft tissue suggesting tenosynovitis of these tendons

Fig. 5.

Post-processing tools on MRI. Maximum intensity projections (MIP) generated from post-contrast images of the hand of a healthy subject (a, b) and a known case of rheumatoid arthritis (c, d). The MIP images show the extent of synovial and tenosynovial involvement with a single image, making it potentially useful in a busy clinical practice

Conclusion

In the evaluation of early RA, the sensitivity of USG was highest for detection of tenosynovitis (86.5%), followed by synovitis (78.6%) and erosions (67.2%). The overall agreement between USG and MRI was found to be more than 80%, and thus USG may serve as a good alternative to MRI for diagnosis of early RA.

Author Contributions

SV, ASB, AKG concepts. SM, SV, ASB UK design. SV, ASB definition of intellectual content. SM, SV, SK literature search. SM, UK, SK clinical studies. SM, SV data acquisition and statistical analysis. SM, SV, ASB data analysis. SM, SV, SK, AKG manuscript preparation. SV, ASB, SK, AKG manuscript editing. SV, ASB, UK, SK manuscript review.

Funding

None.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical standard statement

The study has been approved by the Institute ethics committee.

Informed consent

Informed consent was obtained from all participants as per the protocol of the Institute ethics committee.