Systemic manifestations

Patients with SRV often describe significant weight loss and fatigue, but fever is uncommon.

Musculoskeletal

The patient with SRV typically has changes consistent with long-standing RA, such as joint subluxation and deformity. However, there is often little evidence of an active joint inflammation on clinical examination, but there may be evidence of an inflammation on joint ultrasound.

Cutaneous

Cutaneous involvement has a wide spectrum of features from the minor nailfold lesions to major ulceration with substantial loss of tissue. The development of leg ulcers in a patient with long-standing RA should prompt consideration of vasculitis as the underlying cause; however, other common cause such as venous stasis should also be considered. Features of an ulcer that are suggestive of a vasculitic ulcer include acute onset, pain and location in a position unusual for venous ulceration. The presence of other extra-articular features is suggestive of a vasculitic cause. The differentiation between a vasculitic and non-vasculitic ulcer may, however, be difficult. There is often a reluctance to perform a biopsy because of concerns about poor wound healing.

Digital ischaemia while now very uncommon is indicative of severe ischaemia and needs to be investigated and treated urgently ( Fig. 4 ). Other causes should be sought, such as thromboembolism.

Digital gangrene occurring in a patient with systemic rheumatoid vasculitis.

Nailfold lesions are generally benign ( Fig. 5 ), but should prompt a search for other extra-articular manifestations of RA, in particular interstitial lung disease, scleritis and neuropathy.

Isolated nailfold vasculitis occurring in a patient with rheumatoid arthritis.

Nodules are common and may have overlying vasculitic lesions. They are typically located over points of pressure such as the elbows.

Neurological

Neurological involvement may be sensory or motor or a combination of both. The development of a motor neuropathy should be assessed urgently, as the rapid treatment may prevent progression. The typical pattern is a mononeuritis multiplex with the involvement of arms and legs with a wrist or foot drop ( Fig. 6 ). Sensory neuropathy may be in a glove and stocking distribution suggestive of a length-dependent peripheral neuropathy; this typically affects the legs. It is characterised by burning, parasthesiae and numbness and is not always associated with a motor deficit. The main differential for this pattern of neuropathy is diabetes and alcohol neuropathy.

Mononeuritis multiplex causing wrist drop in a patient with systemic rheumatoid vasculitis.

The central nervous system involvement is uncommon. But patients may present with seizures, neuropsychiatric symptoms, confusional states and haemiparesis. Diagnosis may only be made following meningeal or parenchymal biopsy.

Renal

A clinically significant renal involvement is uncommon in SRV, unlike in other types of vasculitis, such as the ANCA-vasculitides. Presentation is with an active urinary sediment, but progression to renal failure is uncommon.

Ophthalmic

Scleritis and corneal melt are associated with SRV, but may occur independently. The most severe type is corneal melt due to severe peripheral ulcerative keratitis, leading to the thinning of the peripheral cornea and possibly eventually to optic perforation and blindness. It is strongly associated with severe SRV.

Cardiac

The cardiac involvement manifests as pericarditis, aortic incompetence and myocardial infarction. This may be difficult to differentiate from atherosclerotic cardiovascular disease. Symptomatic coronary arteritis is uncommon.

Pulmonary

Pulmonary capillaritis is uncommon in SRV, and other causes of pulmonary haemorrhage should be sought.

Gastrointestinal

Large and small bowel perforation has been described in SRV, but is uncommon.

Radiology

A chest X-ray is needed to find evidence of pulmonary involvement from an interstitial lung disease; if this is abnormal or the patient is dyspnoeic, then pulmonary function tests (looking for a restrictive defect) and a high-resolution computerised tomography (CT) scan may be needed.

Autoantibodies

Patients with SRV are typically positive for rheumatoid factor and ACPA (>90%). Anti-neutrophil cytoplasmic antibodies (ANCA) may be present, typically of a perinuclear pattern or an atypical pattern on immunofluorescence. The specific ELISA assays for anti-proteinase 3 and myeloperoxidase are typically negative. Their presence in a patient with RA does not indicate the development of a separate ANCA-vasculitis, such as granulomatosis with polyangitis (GPA). Antinuclear antibodies are negative.

Complement levels (C3 and C4) are low in SRV, but may also be low in active SLE, hypocomplementaemic urticarial vasculitis (HUVS), mixed essential cryoglobulinaemia, and chronic infections associated with the immune complex formation. It should also be remembered that C3 and C4 are acute-phase proteins and may therefore be elevated in an inflammatory response.

Neurological investigation

Patients presenting with cerebral dysfunction, in association with RA or with evidence of central or peripheral symptoms should be investigated initially with CT or magnetic resonance imaging (MRI) of the brain and spinal cord, as infection, inflammation, malignancy, haemorrhage and ischaemia may all present with neurological symptoms, cerebral vasculitis is uncommon in RA. Many modern RA patients are immunocompromised from the use of methotrexate possibly combined with TNF inhibition or B cell depletion. A lumbar puncture with gram stain, cell count and culture/PCR for bacteria, fungi or viruses is required. Magnetic resonance angiography (MRA) or magnetic resonance venography (MRV) may be necessary if there is a suspicion of a cerebral vasculitis or sinus thrombosis. Peripheral nervous symptoms such as motor or sensory neuropathy may require four-limb nerve conduction studies and possibly a nerve biopsy. The pattern of sensory or motor symptoms can be suggestive of the glove and stocking pattern or mononeuritis multiplex. Electromyography may be required if there is a suspicion of a myopathy.

Tissue biopsy

A tissue biopsy is obviously the gold standard means of confirming the diagnosis of SRV; however, this may not always be possible. In patients with ulcers, there may be concerns about poor wound healing. Ideally, a tissue biopsy should be obtained prior to treatment; however, if there is a critical organ involvement, the treatment should never be delayed for diagnostic purposes. The choice of organ to biopsy is determined by the suspicion of involvement and the ease of biopsy.

A skin biopsy is easy to obtain and has low morbidity, and the samples should be sent for both conventional histochemistry and immunohistochemistry.

Non-biologic immunosuppression

Corticosteroids alone are usually ineffective and rapid changes in dose have been implicated in the pathogenesis and induction of SRV and may predispose patients to vascular occlusion. Azathioprine in combination with prednisolone may be effective in the treatment of SRV, but no controlled data exist . A comparison of azathioprine and prednisolone in the treatment of isolated nailfold vasculitis showed no significant advantage over conventional therapy .

Methotrexate has been used to treat SRV, but there is no evidence from randomised controlled trials to support its use. Other immunosuppressive agents, such as mycophenolate mofetil, have also been used.

The present conventional approach is intravenous or oral cyclophosphamide combined with glucocorticoids, as pioneered by Scott and Bacon in the 1980s, building on the experience with cyclophosphamide in granulomatosis with polyangitis (Wegener’s granulomatosis) The cyclophosphamide dose regimen is similar to that used in the ANCA-vasculitides, that is 10–15 mg/kg IV repeated every 2 weeks for three doses and infusions every three weeks for 6 months . The dose should be reduced to compensate for the effects of age and renal function. The oral dose is 2 mg/kg/day. Glucocorticoids are given IV (methylprednisolone 1 g) with the first dose, and then, a rapid tapering course starting at 1 mg/kg (maximum dose prednisolone 60 mg/day). Patients receiving cyclophosphamide should receive prophylaxis against uroepithelial malignancy with mesna and pneumocystis jirovecii with co-trimoxazole.

Biologic response modifiers

However, this approach is associated with a mortality of 20% at one year, and there is an urgent need for new approaches. The introduction of biologic response modifiers in the treatment of RA TNFi, abatacept, anakinra and B cell depletion has led to considerable interest in their use in SRV. In the Mayo study, cyclophosphamide was used in 29% of the patients and a biologic response modifier in 28%; agents used included TNFi, rituximab, abatacept and anakinra. This heterogeneity in the use of biological response modifiers reflects the lack of clear evidence to support one approach over another . The situation has been further confused by a number of reports of the development of vasculitis in patients receiving TNFi . Indeed, recently Mayo clinic noted that biologic therapy was a risk factor for developing SRV, but this might be a reflection of disease severity rather than being causative.

TNF inhibitors

TNF inhibitors have been used to treat SRV, but were associated with only 56% response rate at 6 months and 9% mortality at 3 years with a high rates of severe infection and relapse .

Rituximab

B cell depletion with Rituximab is now widely used to treat other types of vasculitis, in particular ANCA-vasculitis following the publication of the RITUXVAS and RAVE trials . This use and the experience of using it to treat uncomplicated RA has led to its use to treat SRV. In the French Autoimmunity and Rituximab (AIR) registry, there were 17 cases of SRV treated with rituximab . Thirteen patients were treated with the conventional RA dose of 1 g given twice at a two-week interval, and the remaining 4 received either 0.5 g or 0.75 g at 2 weeks apart or four infusions of 0.75 g and 1 g on four occasions at 1-week intervals. Sixteen patients received concomitant glucocorticoids. Remission was achieved in 6 months in 12 patients and 4 had a partial response and 1 had died. At 1 year, 14/17 patients (82%) were in complete remission as defined by a Birmingham Vasculitis Activity Score (BVAS) of 0 combined with an absence of clinical symptoms and signs of disease. The mean prednisolone dosage was reduced from 19.2 mg/day to 9.7 mg/day, and at 12 months, 14 patients (82%) were in sustained remission. Three patients had severe infections.

The optimum regimen for Rituximab in SRV is unknown, but a logical approach would be to reflect the pattern of use in AAV, with induction using 1 g given twice at an interval of fortnight. The maintenance regimen has not been established in AAV, but current practice suggests that maintenance should be with regular repeat doses of 500 mg–1000 mg every 4–6 months for a total of 2 years .