Key messages

Systemic lupus erythematosus causes hand joint synovitis even in asymptomatic patients and reduces the hand functional strength.

Background

Systemic lupus erythematosus (SLE) is a complex, chronic multisystem autoimmune disease with a variable spectrum of manifestations, ranging from mild musculoskeletal (MSK) manifestations to potentially life-threatening disease [1].

Despite that musculoskeletal (MSK) manifestations have long been thought of mild importance, even patients with no major organ affection raise complaints regarding major functional limitations attributed to MSK affection [2,3,4].

Up to 95% of patients experience arthralgia or arthritis during the course of their disease [5]. Traditionally, SLE arthritis is mainly localized to the small joints, and sometimes, hand deformities resemble rheumatoid arthritis (RA) [6]. Some patients experience limitations in their daily activity due to hand symptoms which may be their only reason for seeking healthcare facilities [3, 7].

Some authors have reported subclinical ultrasonographic (US) synovitis in asymptomatic SLE patients with a high percentage of tendon involvement [8,9,10]. Piga et al. in 2016 [11] found that baseline power Doppler (PD) synovitis score predicted MSK flare within 2 years from US examination. This may help physicians in detecting subclinical joint affection in the early phase of window of opportunity, optimizing medical treatment and suspecting Rhupus patients who display more aggressive MSK involvement.

Joints and tendons US scoring are not standardised yet in SLE [4]. The association between disease activity, MSK involvement, US findings and the functional status of the patient needs to be studied further to improve the patients’ hand function necessary in performing their daily activities. Therefore, we assessed US patterns in SLE patients and determined their relationship with clinical assessment, disease activity and hand functional status using handheld dynamometry.

Methods

This cross-sectional study was conducted on 30 female patients with SLE fulfilling the Systemic Lupus Collaborating Clinics (SLICC) [12] criteria with age more than 18 and less than 65 years old. They were recruited from our Rheumatology and Rehabilitation outpatient clinic serving as a tertiary referral centre over a period of 1 year. Patients were excluded if they were proven to have hand osteoarthritis, previous hand trauma or surgery, and patients with hepatitis C virus-related arthritis, with inflammatory, metabolic or rheumatologic diseases. Patients with peripheral neuropathy, radiculopathy and nerve entrapment neuropathy known to affect the hand function were also excluded following history taking and neurological examination. It is noteworthy that we included 2 patients with Rhupus syndrome.

Patients with hand arthralgia/arthritis on examination date were categorised as symptomatic and those without as asymptomatic [13]. The clinical assessments were done including pain visual analogue scale (VAS) [14,15,16], verbal rating of the habitual intensity of manual labour [17], 28 tender joint count (TJC) and 28 swollen joint count (SJC) [18], presence or absence of hand tenosynovitis [19] and Jaccoud’s arthropathy index (JAI) [20, 21]. Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) was calculated [22]. The mean maximal isometric hand grip and lateral pinch strength were measured using handheld grip and pinch dynamometery (Preston D-704607 PAT) and (Preston D-704608 PAT) respectively. Patients were seated on a chair with shoulders and wrist in neutral position, the elbow in 90° flexion and the forearm parallel to the floor [23, 24]. They bilaterally pressed each dynamometer as hard as possible twice; the higher grip and pinch strength measurements of each hand were recorded and used to calculate the subject’s mean maximal grip and pinch strength. Anti-cyclic citrullinated peptide (ACPA) antibodies were measured using Roche Elecsys immunoassay on Cobas e 411 automated analyser.

Thirty-eight age-matched healthy females were enrolled as a control group for dynamometry testing and intensity of manual labour rating.

Two assessors specialized in MSK US, blinded to the clinical status of the patients, screened the SLE patients according to the European League Against Rheumatism (EULAR) standardized imaging procedures in rheumatology [25] using a high-end US machine, RS80A Samsung Medison, Seoul, Korea, equipped with 3–16 MHz linear transducer (operating at 16 MHz) and PD settings (with avoidance of probe compression). They assessed the 2nd–5th metacarpophalangeal (MCP), 2nd–5th proximal interphalangeal (PIP), wrist joint (radiocarpal and midcarpal joints were assessed as a single unit and scanning was at the radio-lunate joint level) [26], the 2nd, 4th and 6th extensor tendon compartments, and the flexor tendons of the 3rd and 4th digits at the MCP level [27]. Synovial hypertrophy (SH) and PD activity scoring were done using the EULAR-Outcome Measures in Rheumatology (OMERACT) composite PDUS synovitis score (at the joint level) [26, 28]. For group comparison, we calculated the Global EULAR-OMERACT synovitis score (GLOESS) using the previous composite score and the PD index [29, 30]. Each joint was scored for bone erosion and effusion using a binary score (positive/negative), and the number of joints with effusion and erosions per patient was calculated [31]. The total number of pathological tendons or compartments (with grey scale (GS) changes or PD activity) per patient was calculated [27, 32]. It was scored using a semiquantitative score for GS where grade 0 = normal; grade 1 = minimal; grade 2 = moderate; and grade 3 = severe and for PD where grade 0 = no PD signal; grade 1 = minimal (single focal peritendinous signal); grade 2 = moderate (multifocal peritendinous signal); and grade 3 = severe (diffuse peritendinous signal) [27]. We also referred to published consensus-based illustrative images [27]. All lesions detected by GS and PD were confirmed in 2 planes (longitudinal and transverse) [27].

Statistical analysis

Data was analysed by IBM SPSS 20 (Armonk, New York: IBM Corp). Student’s t test or Mann-Whitney test were applied for unpaired continuous variables. The chi-square test with Yates’s correction for continuity or two-tailed Fisher’s exact test were applied when categorial variables/percentages were compared, ANOVA to compare between more than two groups and post hoc test (Tukey) for pairwise comparisons. For bivariate correlations, we used the Spearman rho correlation coefficient. The odds ratio with 95% CI was calculated. P values ≤ 0.05 were considered significant.

Results

This study included 30 SLE patients with a mean age of 40.6 ± 10.3 years and median disease duration of 5.5 (0.16–44) years. Twenty-eight patients (93.3%) received corticosteroids (Cs); the remaining 2 patients (6.7%) were asymptomatic. Hydroxychloroquine, azathioprine and methotrexate were received in 24 (80%), 19 (63.3%) and 6 (20%) patients respectively.

Ninety percent of patients described MSK manifestations at some point across their illness even before establishing the disease diagnosis. History of arthralgia was present in 73.3%, arthritis in 16.7% and morning stiffness (MS) in 40% of patients. Two patients (6.6%) were ACPA positive; only 1 of the 2 (50%) had sonographic erosions. Sixteen patients (53.3%) had clinical tenosynovitis; 6 patients had unilateral hand involvement, while 10 patients were bilaterally involved. Fourteen patients (47%) had joint deformities, of which 9 (64%) had mild deforming arthropathy (JAI ≥ 1 but ≤ 5).

Sixteen patients (53%) were clinically asymptomatic with a median disease duration of 5.5 (0.5–44) years, while 14 (47%) were clinically symptomatic with a median disease duration of 5.5 (0.16–18) years and a median VAS of 30 (20–100) on examination date. There was no statistically significant difference between the 2 groups regarding disease duration (U = 110, MCp = 0.793). Table 1 demonstrates the age, intensity of manual labour and grip and pinch strength in the patient groups and the control group.

Table 1 Distribution of the studied groups according to age, manual labour and grip and pinch dynamometry
Full size table

Table 2 compares between the symptomatic and asymptomatic patients regarding clinical and laboratory findings.

Table 2 Comparison between the symptomatic and asymptomatic patients regarding the different parameters
Full size table

Twenty-one patients (70%) showed US evidence of synovitis with at least 1 joint with SH with a grade ≥ 1 whether active (PD activity) or inactive (no PD activity). It was found in 11 (78.5 %) out of 14 symptomatic patients and 10 (62.5 %) out of 16 asymptomatic. These findings were found to be asymmetric in distribution regardless of hand dominance, where 7 patients (33.3%) had unilateral hand involvement. Eight patients (26.7%) had at least 1 joint with a PD activity ≥ grade 1. Eighteen patients (60%) had ≥ 1 joint effusion, and 6 patients (20%) had erosions. The most frequent US abnormality in the screened joints (540) was SH in 13.3%, while erosions were the least frequent in 1.1%.

Eleven out of 30 patients (36.6%) had US tendon/compartment pathology (25 (8%) out of 300 tendons/compartments). Grade 1 GS was the commonest grading found in 19 (76%) out of 25 pathologic tendons. The 4th extensor compartment was the most frequently affected by GS (7 (28%) out of 25); the 2nd extensor compartment was the most frequently affected by PD (3 (50%) out of 6 tendons with Doppler activity). Grade 1 PD activity was only recorded and limited to the extensor compartments. Table 2 compares between the symptomatic and asymptomatic groups regarding the different US findings. Figures 1 and 2 represent an example of asymptomatic and symptomatic patients respectively.

Fig. 1
figure 1

A 57 years old asymptomatic patient, ACPA negative, anti-dsDNA positive and SLEDAI-2K score of 23. a Left radio-lunate joint with grade 2 SH and effusion. b Right PIP with grade 2 SH and erosion. c Right 6th extensor compartment grade 2 tenosynovitis with grade 1 PD activity

Full size image
Fig. 2
figure 2

A 43 years old symptomatic patient, ACPA and anti-dsDNA negative and SLEDAI-2k score of 16. a Right radio-lunate joint with grade 2 SH. b Right 3rd MCP with erosion. c Left 6th extensor compartment grade 2 tenosynovitis

Full size image

Patients were classified according to both joint deformity and US findings; 8 patients (26.7 %) had non-deforming non-erosive arthritis, 3 (10%) non-deforming erosive arthritis, 9 (30%) mild deforming arthropathy, 1 (3.3 %) deforming erosive arthritis (Rhupus patient), 2 (13.3%) Jaccoud’s arthropathy and 5 (16.7%) had normal joints. Thus, SLE joint involvement in our cohort was 83.3%.

In the total patient population, VAS showed a positive correlation with the number of pathological tendons (rs = 0.48, p = 0.006), duration of MS (rs = 0.5, p = 0.04) and TJC (rs = 0.78, p < 0.0001). JAI had a negative correlation with the intensity of manual labour (rs = − 0.37, p = 0.04) and a positive correlation with the disease duration (rs = 0.44, p = 0.015). The mean grip strength had a negative correlation with the age (rs = − 0.5, p = 0.003) and SLEDAI-2K score (rs = − 0.4, p = 0.02). The mean pinch strength had a negative correlation with the age and the SJC (rs = − 0.45, p = 0.01). The GLOESS had a significant correlation with the cumulative dose of Cs intake (rs = 0.39, p = 0.05). The PD index had a positive correlation with age (rs = 0.48, p = 0.006) and anti-dsDNA level (rs = 0.4, p = 0.02) but had a negative insignificant correlation with the cumulative dose of Cs intake (rs = − 0.029, p = 0.88).

In the asymptomatic group, JAI showed a positive correlation with the number of joints with sonographic effusion (rs = 0.67, p = 0.004) and a negative correlation with the mean pinch strength (rs = − 0.49, p = 0.050). The PD index had a negative correlation with both the mean maximal grip and pinch strength (rs = − 0.5, p = 0.04 and rs = − 0.6, p = 0.01 respectively). The number of joints with erosions correlated with the number of pathological tendons/ compartments (rs = 0.51, p = 0.04).

In the symptomatic group, the number of joints with erosions correlated with SJC (rs = 0.61, p = 0.02).

Discussion

The magnitude of MSK involvement

In this study, different degrees of joint involvement have been shown to be underestimated in SLE patients, where 83.3% of patients had either clinical or sonographic joint involvement. Both symptomatic and asymptomatic patients had comparable US scores with no significant difference between them. Ten (62.5%) out of 16 asymptomatic patients had a sonographic evidence of synovitis. This finding has recently been described in the literature by “subclinical synovitis” [8, 9, 13, 33]. No significant difference was found between both patient groups regarding PD index probably because they had comparable cumulative dose of Cs intake. Ruano et al. [8] suggested that the high prevalence of steroid use among the SLE population could reduce the global inflammatory burden and result in lower prevalence and grading of the US PD findings.

In addition, 16 patients displayed clinical tenosynovitis, while sonographic tendon/compartment involvement was only found in 11 patients. The discrepancy between the clinical and sonographic data has been attributed in literature to the presence of pain related to depression, fatigue and neuropathic pain with central sensitization or overlap with fibromyalgia leading to an exaggerated symptomatology [34, 35]. Lins and Santiago [36] and Mosca et al. [37] agreed that US cannot be used as a confirmatory diagnostic method of symptomatology since not all clinical changes necessarily correspond to anatomic lesion in US (synovitis/tenosynovitis).

Despite the magnitude of joint and tendon involvement, low GS and PD scoring was the most frequent sonographic finding in both joints and tendons, explaining the subclinical presentation. It seems that despite the devastating major organ damage in SLE, MSK affection is the least worrisome but not the least frequent. The weight of SLE-related joint damage in comparison with other organ damage is 1 out of 47 according to the SLICC/American College of Rheumatology Damage Index [38]. However, arthralgia is as frequent as 85–94% [39, 40]. In our study, 21 patients (33.3%) had unilateral hand joint involvement which disagrees with the previous understanding that SLE displays a symmetrical joint affection [5, 21].

Erosions were the least frequent sonographic abnormality, and they were all detected in the dominant hand suggesting a mechanical causative factor. This could be supported by the absence of localized joint tenderness at erosion sites and its correlation with the number of tendons with sonographic abnormality. Inflamed tendons, lax ligaments, and pulleys cause mechanical disadvantage leading to erosion formation. Van Vugt et al. [41] suggested that bone erosions in SLE, other than Rhupus, are caused by friction rub of the overlying inflamed tendons or ligaments. The clinical impact of erosions remains undetermined since no correlations were found with either deformities (JAI) or hand strength.

Are there limitations in hand function?

Grip and pinch strength were significantly impaired in both asymptomatic and symptomatic patients compared to the control group. The mean grip strength had a significant negative correlation with SLEDAI-2K score denoting that disease activity is an important determinant of the grip strength necessary in performing activities of daily living. In addition, we also found a negative correlation between the PD index (marker of active synovitis) with both mean grip and pinch strength in the asymptomatic group. In a study of RA patients, Dedeoğlu et al. [42] found a similar correlation between the disease activity and both grip and pinch strength. In addition, the lateral pinch strength negatively correlated with JAI, while the grip did not. The pinch requires more precision and isolation of the thumb and the index, making it more liable to the extent of joint deformity and active synovitis. Vlieland et al. [43] suggested that flexion attitude or deformity of the thumb and MCP subluxation consistently attribute to impaired hand function in RA patients.

It seems that the deficit in grip strength due to joint inflammation or deformity in SLE was compensated by the unaffected surrounding joints contributing to the grip strength. However, upon pinch strength testing, the deficit was unmasked.

What determines deformities?

Forty-seven percent of the studied patients had various grades of deformities and JAI positively correlated with disease duration, which is in accordance with the published literature [41, 44, 45]. Also, a significant negative correlation was found with intensity of manual labour; the more intense the labour, the less liability for deformity formation. The deformity in SLE seems to be the consequence of ligament laxity combined with muscle imbalance, rather than the destructive effect of synovitis as in RA [41]. It seems that intense manual work guards against muscle wasting with subsequent deformity prevention and preserves joint mobility.

In the asymptomatic group, JAI correlated with the number of joints with sonographic effusion while it did not correlate with the GLOESS, PD index or tendons pathology. We also suggest that joint effusion is attributed to the mechanical disadvantage caused by the deformities rather than an intrinsic joint pathology. This fact is supported by the recent OMERACT criteria which disregarded effusion as a sign of synovitis [28].

What caused hand arthralgia?

The VAS did not correlate with either the GLOESS or the joints PD index, while it correlated with the number of pathological tendons. Torrente-Segarra et al. [13] found a higher frequency of tenosynovitis (39.2%) in SLE hand arthralgia group compared to synovial hypertrophy (25%) or active arthritis (14.2%). In addition, Doppler activity was significantly higher in their arthralgia group compared to the non-arthralgia group, which contradicts our findings, probably due to the summation of the tendons and joints Doppler status.

Limitations

Prospective longitudinal studies are necessary to explore the outcomes of subclinical joint affection and silent erosions. In addition, the impact of hand isometric strengthening exercises and patient education programs concerned with joint protection on grip function and deformity prevention needs to be investigated.

Conclusion

SLE patients may have higher subclinical synovitis, erosions and tendon involvement than expected, which may in turn reduce hand grip and pinch strength. Disease activity may also have a negative impact on the hand grip functional strength. The asymmetric pattern of US findings irrespective of hand dominance suggests the importance of performing bilateral joints screening, with priority to the wrist and MCPs.