Keywords

Systemic lupus erythematosus (SLE) is a chronic immune-mediated disease affecting multiple organs including skin, brain, peripheral nervous system, heart, gastrointestinal tract, kidney, and, almost invariably, joints. Clinical features in individual patients are highly variable, and arthritis and arthralgias in SLE deserve to be specifically addressed.

1 Arthritis and Arthralgias in SLE

Articular involvement is very frequent in SLE. In 34–50 % of patients with SLE, joint involvement is the first manifestation of the disease, whereas during the course of SLE, it is almost always present [1–4]. Joint involvement occurs more frequently in women, both at onset and during the course of the disease [5–7]. It is polymorphous and shows varying degrees of clinical severity, with increasing frequency depending on age at SLE onset, and it is more frequent in patients with onset of illness in adulthood [8, 9]. It represents one of the most frequent causes of difficulties in daily activities, job reduction, and abandonment [10–12].

Articular signs, such as arthralgia or arthritis, may be observed in the course of SLE. Arthralgia, which is defined as the presence of joint pain in the absence of clear synovitis, erosion, or deformities, is very frequent both at the onset and during the course of the disease [13]. Arthralgia is persistent, migrant, transient, and frequently associated with myalgia. It is often extremely intense and disproportionate to the finding at the physical examination [14]. In patients with arthralgia alone, the most recent imaging techniques, such as magnetic resonance imaging (MRI) and ultrasonography (US), may frequently reveal inflammatory articular changes which is undetectable by clinical evaluation [15–17].

Arthritis in the course of SLE can be acute, subacute, and/or chronic and is often detected (in 67–87 % of cases) (Table 6.1) at disease onset. The differences in prevalence reported in various studies may depend on the racial backgrounds of the cohorts, on the different age of the patients, and on the level of expertise of the enrolling center (e.g., nephrology centers or rheumatology centers).

Table 6.1 Prevalence (%) of arthritis in different series of adult patients of SLE
Full size table

Acute arthritis may occur as polyarthritis or oligoarthritis; it may be symmetric or asymmetric, with preferential involvement of the joints of the hands, wrists, and knees [18–26]. It is sensitive to anti-inflammatory treatment and usually does not recur during therapy [14]. It is frequently associated with visceral involvement and systemic signs; thus it can be a warning symptom of an SLE flare.

Subacute arthritis has a more prolonged course with milder inflammatory signs and it is frequently accompanied by morning stiffness. In older patients, the clinical picture may resemble rheumatic polymyalgia [14]. Lastly, arthritis in the course of SLE can occur as chronic arthritis without deformities and erosions, or as deforming, non-erosive, reversible arthropathy, or in some patients as erosive arthritis.

Examination of the synovial fluid shows some inflammatory liquid with a predominance of mononuclear cells and a decrease in complement levels with hypergammaglobulinemia. Antinuclear antibodies (ANA), lupus erythematosus (LE) cells, hematoxylin bodies, and ragocytes may be observed [14]. In a recent study, LE cells were present in 5/31 patients suffering from SLE and in 9/27 patients with the overlap syndrome (rheumatoid arthritis (RA)/SLE). In the same study, LE cells were observed in 2.6 % of 331 patients with RA and in none of 4 subjects with Still’s disease, in 9 with systemic scleroderma, in 132 with ankylosing spondylitis, in 57 with Reiter’s syndrome, and in 34 with psoriatic arthritis [27].

Common findings at synovial biopsy include synoviocyte hyperplasia, scarce inflammatory infiltrate, vascular proliferation, edema and congestion, fibrinoid necrosis and intimal fibrous hyperplasia of blood vessels, presence of fibrin on the synovial surface, and fibrin-like deposits in the chorion [28]. Indeed, some synovial alterations, ranging from simple hyperemia to synovitis, may mimic those found in RA [14]. Some studies have found differences in the synovium of patients with SLE and those suffering from RA and osteoarthritis (OA) from a gene expression profile perspective [29]. The synovium of patients with SLE suffering from arthritis shows a very distinct molecular signature as compared to what is observed in patients with OA and RA. It is characterized by the upregulation of interferon-inducible genes, as observed in the peripheral blood and kidney glomeruli of SLE patients [30], and downregulation of genes involved in extracellular matrix homeostasis. This might suggest the presence of different pathogenic mechanisms in SLE and RA, which would explain the lack of bone erosion observed in SLE patients with arthritis [29, 31].

2 Jaccoud’s Arthropathy

The 1982 ACR classification criteria [32] modified the weight of joint involvement switching from “arthritis without deformity” of 1971 criteria [33] to “non-erosive arthritis.” This change allowed to include patients with Jaccoud’s arthropathy (JA) among those with SLE. In 2012, the SLICC criteria [34] proposed modifying the criterion to “synovitis ≥2 peripheral joints, characterized by pain, tenderness, swelling or morning stiffness ≥30 min,” since new imaging techniques clearly show that some forms of SLE arthritis are in fact erosive [17].

In 1869, Jaccoud [35] first described deforming, non-erosive, reversible arthropathy associated with rheumatic fever. Later, this arthropathy was also observed in other rheumatic diseases and connective tissue diseases, as well as in sarcoidosis, infections, hypocomplementemic urticarial vasculitis, chronic pulmonary disease, inflammatory intestinal disease, pyrophosphate deposition disease, hypermobility syndrome, borreliosis, and neoplasia [36–41]. JA also occurs in an idiopathic form, in particular in the elderly, sometimes affecting several members of the same family [42, 43]. It was first described [44, 45] in patients suffering from SLE [24, 46–48] with a prevalence of up to 35 % [24, 26, 47–52].

The articular deformities of JA may be limited to the ulnar deviation at the metacarpal-phalangeal joint (MCP). Villaumey [53] and Alarcon-Segovia [46] considered this a diagnostic element for JA in the absence of erosions and rheumatoid factor (RF). JA may be widespread and can simulate evolved RA with lateral hyperlaxity of the distal interphalangeal articulations, swan neck deformities, boutonniere deformities, Z-shaped thumb, and carpal hyperlaxity. JA may also affect the feet [54, 55] and knees, sometimes in a disabling way, as well as the shoulders [14]. Deformities are usually reducible, even though some deformities may present elements of fixity [13].

In 1992, Spronk and coworkers [47] proposed a diagnostic index for JA (DIJA) based on the presence or absence of ulnar drift (>20′), swan neck deformities, boutonniere deformities, Z deformity, and limited MCP extension. Depending on the number of affected fingers, the first four items of the diagnostic index are graded from 2 to 3, whereas the fifth item is graded from 1 to 2, with a maximum total score of 23. JA was considered as being present if the index exceeded five points. This scoring method was frequently cited but never validated in well-designed studies [13, 56]. In 1998, van Vugt [57] introduced the term “mild deforming arthropathy” for patients who present DIJA with a score equal to or below 5 in the presence of deforming arthropathy without erosion. New imaging techniques led to the detection of erosive signs not previously detected by conventional radiology, thus leading to a new step in the differentiation between JA, mild deforming arthropathy, and erosive arthritis [26]. Recently, new criteria to differentiate between “idiopathic” and “senescent” JA were proposed by Santiago et al. [56], including: (1) typical joint deformities which are correctable in a passive position, (2) presence or history of articular inflammation in the deformed joints, regardless of its intensity or etiology (RA, SLE, etc.), (3) absence of similar deformities in other healthy members of the same family, and (4) no erosions on conventional radiology, magnetic resonance, or high-performance ultrasound examination.

The presence of JA in patients with SLE has been associated with older age [58] and disease duration [47, 59, 60]. However, these findings were not in agreement with the observations by Alarcon-Segovia et al. [46]. Besides it has been shown that the main determinant for JA was high disease activity in the absence of synovitis [61].

JA is positively associated with Sjogren’s syndrome [46, 58] and frequent tendon rupture [62] and negatively associated with renal involvement [24, 51].

Some authors have pointed out an association with the presence of RF [46, 47], but this finding was not confirmed by other studies [51, 59]. JA was associated with higher levels of C-reactive protein (CRP) [47, 60, 63], with the presence of lupus anticoagulant and anticardiolipin antibodies [57], and with the presence of antibodies against U1 RNP [64], and inconstantly with anti SS-A/Ro and -B/La [47, 51, 58, 65, 66]. A correlation with the presence of antibodies to type II collagen [67], with higher levels of interleukin-6 (IL-6) [60] and with anti-double-stranded DNA antibodies [46], was observed. Interestingly, JA has never been associated with antipeptide citrulline (anti-CCP) antibodies [68, 69].

In general, the development of JA is correlated with abnormalities of soft tissues, with ligament and capsular laxity, relaxation, and subsequent deviation of tendons from their axis with the association of muscular dysfunction. Some authors have speculated that the laxity of the articular capsules and ligaments may be secondary to inflammation with fibrosis of the articular capsule [70, 71]. A role of synovial vasculitis [57] has been proposed. Besides, it has been also hypothesized a role in JA development for a synovial vasculitis [57], persistent inflammatory process of the synovium with inflammatory cells infiltrate and IL-1 and IL-6 production [47, 72]. These observations are in line with the high levels of CRP observed in patients with JA [47, 60, 63], but not in other patients with SLE even during disease flare [26].

The detection of high levels of RF in patients with JA is inconsistent [46, 47, 51, 59]. Thus, the presence of RF may act as a local inductor of the inflammatory process through the formation of immune complexes. However, it is worth noting that despite the documented presence of synovitis, this is not as aggressive as what is observed in RA [56]. JA can appear in other conditions as dermatomyositis, scleroderma, and chronic pulmonary disease without evidence of previous arthritis [40]. In 2006, Caznoch et al. [50] suggested a possible role of hyperparathyroidism linked to renal failure and the presence of an association between JA and the hypermobility syndrome, contrasting however with the previous observations by Klemp et al. [73]. A possible role for tenosynovitis has also been hypothesized [40], considering the reported association between JA and tendon ruptures [62]. As a matter of fact, 26 % of the 55 tendon ruptures in patients suffering from SLE were associated with the presence of JA [62]. Histological reports are very limited, though they led to the detection of mild synovitis without significant proliferation of the synovial membrane, light inflammatory infiltrates, microvascular alterations, fibrin precipitates, and hematoxylin corpuscles [47].

Based on a previous radiologic description, deforming chronic arthritis may present with swelling of the soft tissues; juxta-articular osteoporosis and joint space narrowing are rare, and exceptionally hook shaped erosions in the hands (and feet). All these features differ from what observed in RA (i.e., damage to the radial site of the metacarpal heads as well as a well-defined hook-shaped deformity with a sclerotic margin that is considered an adaptation to local stresses of persistent ulnar deviation) [74].

3 Rhupus

During the course of SLE, it is possible to observe, albeit rarely, an erosive arthritis similar to what is observed in RA. In 1971, Peter Schur [75] coined the term “rhupus” to describe patients with SLE who present arthritis and who also fulfill the classification criteria for RA [76]. Currently, the term rhupus is used by some to describe the coexistence of SLE and RA in the same patient [49, 57], while others use it to outline a subset of SLE patients with distinctive articular signs and typical clinical and radiological characteristics [3, 77]. However, the definition is still disputed since the immunopathological processes of SLE are considered to be exactly the opposite of the RA processes [78]. The real prevalence, the natural history and the clinical appearance are supported by few case series and small cohorts of patients, though with discrepancies in the definition of the cases and assessment methods [3, 77, 79–82]. It is however a very rare variant considering that until 2013, only 150 cases had been published [83]. Recently, the number of reports has been enriched by additional cohorts of patients [49, 84–86]. An epidemiological study showed a prevalence of about 0.09 % [79]. The prevalence of the more recent studies ranges from 1.30 to 5.8 % [49, 84, 85], which is surely higher than in previous reports [3, 79]. New imaging techniques for the study of the musculoskeletal system, like MRI and US [15–17, 70, 87, 88], have allowed to detect erosive alterations that would otherwise be undetectable by conventional radiology and to stress the higher prevalence of rhupus (9.7 %) [84].

Clinically, there are no statistically significant differences between the rhupus group and the control group regarding age and sex [49, 84, 85]. The signs of RA usually precede those of SLE [49, 84, 85] and these subjects have slower disease progression compared to patients with SLE without articular erosions [3, 49, 84, 89]. There were no significant differences in the prevalence of anti-double-stranded DNA, anti-Sm antibodies, anti- nuclear antibodies, and antiphospholipid antibodies between rhupus and SLE patients [49, 84]. On the other hand, patients suffering from rhupus present increased erythrocyte sediment rates and CRP levels [49, 84] and a greater presence of RF and antipeptide citrulline (anti-CCP) antibodies [49, 68, 84, 86, 90–94]. With regard to the presence of anti-CCP antibodies, a recent meta-analysis conducted on seven studies revealed 91.8 % and 47.8 % pooled specificity and sensitivity, respectively, in erosive arthropathy in SLE [92]. Comparatively in RA, a meta-analysis found the specificity and sensitivity of anti-CCP antibodies to be 95 % and 67 %, respectively [95]. The pooled specificity of anti-CCP antibodies in SLE patients with erosive disease is slightly lower than that of RA. It must be stressed that the studies by Qing et al. [93] and Zhao et al. [94] showed the highest sensitivity and lowest specificity, while Budhram et al. [92] hypothesized that these results can be explained by a threshold effect due to the low cutoff of anti-CCP antibody positivity that was used by the authors. Five studies reported anti-CCP antibody positivity in only 5 % or fewer patients with SLE and non-erosive arthritis [68, 91, 96–98]. The meta-analysis by Budhram suggests that the specificity of anti-CCP antibodies in SLE patients with erosive disease is comparable to that of RA when high cutoffs are used [92].

In rhupus, there is also a correlation between the severity of the articular involvement and positivity for anti-CCP antibodies [90–94]. A more recent study again confirms the association between anti-CCP antibodies and rhupus and highlights that the presence of these antibodies in patients suffering from SLE must make clinicians aware of the coexistence of RA [86], as confirmed by others Authors [69, 92].

Patients with rhupus have mild SLE activity and a lower incidence of visceral organ involvement compared to patients with SLE without RA and in particular with regard to renal and neurological involvement [49, 81, 84], although important clinical manifestations have also been reported [99, 100]. Patients with rhupus show a predominance of manifestations that are typical of RA, including clinical inflammation, deformities and erosions, and rheumatoid nodules, as well as a significantly high prevalence of RF and anti-CCP antibodies [3, 49, 69, 79, 81]. High titers of RF and anti-CCP antibodies are very often observed in patients meeting ACR criteria for RA and presenting articular erosions [69, 92]. Some authors believe that the appearance of rheumatoid nodules in patients with SLE represents a risk factor for rhupus [101].

A review of the medical literature on the correlation between anti-CCP antibodies and erosive arthritis in patients with SLE led Budhram et al. [92] to hypothesize the presence of two subgroups of patients with erosive arthritis. One subgroup presents a process that is pathologically different from that of RA, their anti-CCP antibodies are often negative, and these subjects likely do not meet the ACR criteria for RA. In the second subgroup, the pathogenesis of the erosions is the same as in RA, the anti-CCP antibodies are often positive, and these subjects likely meet the criteria for RA. According to Budhram et al., patients in the two groups may present different erosive manifestations; in the former they may be similar to those described by Pastershank in JA [74], while the latter may show marginal erosions that are typical of RA. This theory should be investigated with specific imaging studies since the new imaging techniques, such as US and MRI, have shown an unexpectedly high prevalence of subclinical articular and periarticular involvement and, most importantly, a higher prevalence of bone erosions as compared to standard radiography, as well as an unexpectedly high prevalence of synovitis and tenosynovitis [15, 17, 102–104]. This hypothesis is supported by the observations that anti-CCP antibodies and RF show an additive effect on erosion and erosion size and number in RA [105].

Alongside the arthritic manifestations, tenosynovitis is frequently found in patients with SLE, mainly localized to the extensor tendons in the hands, with a prevalence ranging from 28 to 61 % [15, 17, 102–104].

4 Joint Involvment Management

There are some guidelines regarding the treatment of arthritis in the course of SLE [106]. An approach based on the type of arthritis has been proposed [107, 108]. First-line, short-term treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) is frequent in the presence of acute or subacute arthritis, given the episodic and limited nature of articular flares in many patients with SLE [107, 108]. Clearly, administration of these drugs must include rigorous renal and hepatic monitoring; cardiovascular and skin photosensitivity risks must also be taken into consideration, as should reports on rare cases of aseptic meningitis [108]. Many patients cannot tolerate NSAIDs or may present contraindications to their use, while others show more persistent arthritic episodes that are refractory to these drugs. In these cases, antimalarial drugs are recommended, in particular hydroxychloroquine, usually at a dosage of 6 mg/kg/die [109]; if necessary, corticosteroids should be associated [107–110]. Low-dose corticosteroids (≤10 mg/day of prednisone equivalent) are usually administered, and if high doses are needed, corticoid-sparing agents should be used [108]. Direct injection of corticosteroids into joints can be useful, especially when involvement is limited to one or few joints or in tenosynovitis. In patients who do not respond to corticosteroids or antimalarial drugs, methotrexate or other immunosuppressants (azathioprine, leflunomide, mycophenolate mofetil) or non-conventional therapies must be taken into consideration [107, 111–113]. There is broad consensus on the use of methotrexate as a steroid-sparing agent [111] and its effectiveness in controlling articular manifestations [110, 111, 114–118]. Leflunomide proved to be effective in a controlled study of only 12 patients [119]; however, several cases of cutaneous lupus flare-up were reported [108]. Mycophenolate mofetil was also assumed to be effective [112, 120–122], even though evidence in the literature is limited [108]. Among the non-conventional drugs, some reports indicate the efficacy of rituximab [112, 122–124], even if there are no specific controlled studies on arthritic manifestations in SLE. With regard to belimumab, literature data [121, 126–128] suggest that it has potential for use in the treatment of severe joint symptoms in lupus that are resistant to corticosteroid treatment or refractory to conventional treatment [108, 113], whereas tocilizumab [129] is still being investigated.

Despite its reduced efficacy and frequently unfavorable effects [130], if all other therapies fail, abatacept may be taken into consideration (under strict control of a reference center or other experts) for patients in whom lupus manifests as a corticodependent joint disease [108], while the use of anti-TNF-alpha drugs is not recommended [108, 125].

Treatment of Jaccoud’s arthropathy is the same as what is recommended for chronic arthritis in the course of SLE. However, despite their symptomatic effect, there are no guarantees that NSAIDs, low-dose steroids, antimalarial drugs, or methotrexate can inhibit the progression of deformities [56]. The benefits of physical therapy and the use of orthotic devices are yet to be demonstrated, and there are few reports on surgical procedures to correct JA [131–135] since the indications, the best modalities, and when to indicate them are still unknown [50].

There are several reports concerning the use of disease-modifying antirheumatic drugs (DMARDs), and in particular methotrexate, in rhupus [117, 136, 137], even in combination with other drugs [138]; however, many of them showed inadequate response [26]. There are small case series reporting the efficacy of mycophenolate mofetil [139] and cyclosporine [140]. Anti-TNF-alpha, which is effective in treating RA, seems to be less effective in treating rhupus and SLE. It can induce the production of autoantibodies, such as antinuclear antibodies and anti-DNA antibodies [141], and more rarely it may result in lupus manifestations in both RA [142, 143] and rhupus patients [144]. Thus, the use of anti-TNF-alpha is not recommended in SLE patients [125]. The use of rituximab shows more encouraging results, as demonstrated in small clinical series in open clinical studies [83, 145] as does abatacept [130, 146]. Finally, two of our patients benefited from the use of tocilizumab [147].

There is a need for further controlled studies and, consequently, specific guidelines for the various forms of arthritis during the course of SLE.