The British Society for Rheumatology Guideline for the Management of Systemic Lupus Erythematosus in Adults

Caroline Gordon; Maame-Boatemaa Amissah-Arthur; Mary Gayed; Sue Brown; Ian N. Bruce; David D'Cruz; Benjamin Empson; Bridget Griffiths; David Jayne; Munther Khamashta; Liz Lightstone; Peter Norton; Yvonne Norton; Karen Schreiber; David Isenberg


Rheumatology. 2018;57(1):e1-e45. 

In This Article


Overview of Treatment of Mild Lupus

Mild lupus features (Table 7) are distressing for patients and warrant treatment to relieve symptoms and signs. Such treatment may prevent progression to severe manifestations requiring more intense immunosuppression. These manifestations can be managed with CSs, HCQ and other antimalarials, MTX, NSAIDs and sunscreens. The LOEs and GORs for the drugs used to treat lupus disease are summarized in Table 2, and the SOAs with the recommendations are above. There are little data to support the use of topical therapies, dapsone, retinoids, thalidomide or danazol in the treatment of refractory cutaneous lupus rashes and vasculitis, and as these drugs are not used for other systemic features of lupus, they are not discussed here but have been reviewed.[287,288]

CSs for Mild Lupus

Summary. Topical preparations should be used initially for cutaneous manifestations, and intra-articular (IA) or intramuscular (i.m.) injections of CSs for arthritis. Short courses of oral prednisolone (up to 20 mg/day) are used for short periods of time (up to 14 days and reduced rapidly) to induce remission in some cases of mild lupus where local treatment is not sufficient or practical (evidence discussed below in moderate lupus). Prednisolone can be used in women who are trying to conceive, are pregnant or are breast-feeding.[239]

Evidence. There are no RCTs comparing different types of CS administration, such as skin creams and ointments, intralesional, IA and i.m. injections, and oral CS drugs (usually prednisolone in the UK). CSs contribute to the development of chronic damage and co-morbidities such as cataracts, osteoporotic fractures, diabetes, atherosclerosis and infection.[12,14] It has been shown that a 1 mg/day increase in maintenance prednisone dose is associated with a 2.8% increase in the risk of new organ damage, and that prednisolone dosing of ⩽7.5 mg/day is associated with less risk of cataracts, osteoporotic fractures and cardiovascular damage than higher doses.[306]

Conclusions. The lowest possible dose/amount of CSs should be used due to their side effects, including the risk of contributing to chronic damage and infection. Prednisolone treatment at a low dose of ⩽7.5 mg/day may be required for maintenance therapy and has less risk of side effects than higher doses (2+/C).

HCQ and Other Anti-malarial Agents

Summary. There is good evidence (Table 2) for the efficacy and safety of HCQ, the most commonly prescribed anti-malarial agent and one of the few licensed drugs for lupus. Providing that the patient has normal renal and liver function, HCQ can be used at doses of up to 6.5 mg/kg/day and is compatible with pregnancy and breast-feeding. It is used (Table 7) for skin and joint involvement, myalgia, fever, fatigue, pleurisy, to reduce the development of renal disease and chronic damage[14,121] and for its steroid-sparing properties (even in patients with more severe disease).[71] Chloroquine is used if HCQ is not available or not tolerated; however, there is less evidence for benefit and it has a greater risk of retinal toxicity than HCQ.[121] Mepacrine (quinacrine) is used predominantly for cutaneous lupus and has the least risk of ocular toxicity.[287,307–309]

Evidence. The benefits of anti-malarials on lupus activity were reported in four RCTs,[81–84] five prospective cohort studies,[87–91] three retrospective cohort studies[92–94] and an open-label extension of the first RCT.[95] There have been two other double-blind RCTs confirming that lupus rashes significantly improve with HCQ[85] and chloroquine.[86] The cohort studies have shown that response often takes 3–4 months,[94] but at 6 months only 60% of patients with discoid rash show some response.[94] Another study showed that 20% of patients with an adequate response lose it within 2 years and need other therapies.[310] Higher drug levels were associated with increased cutaneous response in a prospective study.[311] In a double-blind RCT,[80] low drug levels were associated with increased disease activity. Systemic features and smoking are also associated with an increased risk of poor response.[94,96,122]

Many of the studies showing increased flare rates in patients who discontinued HCQ involved pregnant patients. A RCT in lupus patients[84] and two prospective[87,90] cohort studies support the use of this drug before conception and in pregnancy to reduce flares in the mother. Although HCQ can cross the placenta, exposure is not associated with significant adverse effects on the fetus.[87,90,97–100] HCQ has anti-thrombotic as well as anti-inflammatory properties and by reducing disease activity in the mother may improve the outcome for the child by improving placental function.[101,102] There is increasing evidence that HCQ reduces the risk of CHB in babies born to mothers with anti-Ro antibodies.[103,312,313] Further evidence supporting the use of HCQ in pregnant women as well as in those planning pregnancy and breast-feeding is reviewed in the BSR Guidelines on drugs in pregnancy in the rheumatic diseases.[239]

There is further evidence from high-quality prospective and retrospective cohort studies that patients treated with anti-malarials (particularly HCQ) not only have lower levels of overall lupus activity and reduced rates of flare,[80,81,84,89,90,95] but can be managed with lower doses of CSs.[83,84,90,104] The patients are more likely to stay clinically quiescent if HCQ is continued when the disease goes in to remission.[105] Patients on MMF are more likely to achieve renal remission if treated with HCQ.[93] Patients on HCQ are less likely to develop serious renal disease and have delayed time to renal damage,[104] lower frequency of seizures[106] and less NP damage,[107] greater delay in integument damage,[108] less overall damage[109,110] and, most importantly, improved survival.[111,112] Some of the benefits on survival may be mediated by the beneficial effects of anti-malarials on total cholesterol, LDL-cholesterol, triglycerides, glucose[113] and/or by the prevention of thrombosis[101,102,121] and atherosclerotic plaque formation.[114]

Patients take HCQ on average for about 6 years.[115–118] In general HCQ is well tolerated and better tolerated than chloroquine.[86,115,116,121] The commonest adverse effects of anti-malarials are gastrointestinal, but a few patients stop because of headache, dizziness, itching, rash, non-retinal eye problems, hearing loss, myopathy or other rare neuromuscular side effects.[121,287] The most serious adverse events are cardiac (which are very rare)[119] and retinopathy (which is more common with chloroquine than HCQ).[121,314] Retinopathy is unpredictable but unlikely with <7 years treatment with HCQ. It is more common thereafter[120] and with doses of HCQ above 6.5 mg/kg/day, or renal or liver impairment. It requires active screening to detect it early when it is asymptomatic and is most likely to be reversible.[120,314] Policies on screening for ocular toxicity vary between countries and local guidelines should be followed.[314,315] In general in the UK, baseline and yearly optician eye tests are recommended initially, with more detailed ophthalmological screening after 5 years of therapy.[316]

Conclusions. There are good data from two systematic reviews and a meta-analysis including 7 RCTs and 36 cohort studies supporting the use of HCQ in lupus patients to reduce disease activity and as a steroid-sparing agent: overall LOE 1 ++, GOR A. HCQ should be given to all patients with mild lupus to prevent flares, the development of damage and to improve survival. It is recommended that HCQ be continued or started, even in those developing disease severe enough to warrant immunosuppressive therapies, including LN.[22,24,25] However patients with renal or liver dysfunction should have the dose reduced.[314] It is compatible with conception, pregnancy and breast-feeding. Unfortunately, it has a long half-life and takes at least 2 months to be effective.[287,309] Patients need to be warned about this or they may discontinue the drug prematurely.

MTX in Mild SLE

Summary. Although not licensed for the treatment of lupus, low-dose weekly MTX (⩽25 mg/week) has been used to reduce mild and moderate disease activity in lupus, particularly to control inflammatory arthritis and lupus skin rashes, originally on the basis of a variety of case series and cohort studies.[317,318] MTX was originally used in patients who had failed HCQ and low-dose CSs, but it can be used with HCQ to avoid CSs or to promote CS dose reduction. Caution has been advised on the use of MTX in patients with LN, particularly as those with renal impairment will be at increased risk of MTX toxicity.[317] It is contra-indicated in women trying to conceive or pregnant as it is teratogenic. For these patients AZA would be more suitable (see section on moderate lupus for evidence).

Evidence. A systematic review by Sakthiswary and Suresh[319] summarizes the data from three controlled trials (two double-blind, placebo-controlled trials,[123,124] and a controlled open-label trial comparing MTX and chloroquine[125]) and five observational studies (two open-label prospective studies;[126,127] a cross-sectional study;[128] a retrospective case–control cohort study;[129] and an open-label controlled study[130]). Another systematic review[133] includes two additional case series.[131,132] These studies support the use of MTX to reduce mild and moderate lupus disease activity, and some demonstrated steroid-sparing properties. Some of these studies showed benefit specifically in treating lupus arthritis, rashes, vasculitis, serositis, myositis and constitutional symptoms, but there was little change in ESR, anti-dsDNA antibodies, C3 or C4 levels, except in a study with longer duration than previous studies.[130] The reduction in SLEDAI in the five controlled studies reporting these data included in the systematic review[319] was calculated to have an odds ratio = 0.444 (95% CI: 0.279, 0.707; P = 0.001). The analysis of the four controlled studies reporting steroid-sparing properties for MTX provided an odds ratio = 0.335 (95% CI: 0.202, 0.558; P = 0.001). Side effects led to discontinuation in ~10% of patients but were not serious. It is teratogenic and should not be used in women within 3 months of planning to conceive, or who are pregnant or breast-feeding,[239] nor in patients with renal impairment, because reduced renal function increases the risk of adverse events, particularly bone marrow suppression.

Conclusions. There are good data from two systematic reviews including three RCTs and seven cohort studies supporting the use of MTX in lupus to reduce disease activity and as a steroid-sparing agent: overall LOE 1+, GOR A.

NSAIDs in Mild SLE

Summary. There are no RCTs of NSAIDs in SLE. Publications support the cautious use of NSAIDs for short periods of time for symptom control in SLE (inflammatory arthralgia, myalgia, chest pain and fever) where potential benefit outweighs the known risks of NSAIDs and paracetamol has been insufficient or not tolerated. The risk of NSAID-induced acute renal failure is increased in patients with LN, so NSAIDs should be avoided in patients with renal involvement. NSAID-induced allergic reactions, aseptic meningitis, cutaneous reactions and hepatotoxicity are increased in SLE patients. Caution is required in pregnancy.[240]

Evidence. A review of the literature on non-selective Cox inhibitors and selective Cox-2 inhibitors[320] highlighted the potential increased risk of renal, hepatic and neurological toxicity in lupus patients. A retrospective case series assessing celecoxib, with a detailed literature review of NSAIDs[321] and a more comprehensive systematic review addressing the risk–benefit ratio of non-selective and selective inhibitors of cyclooxygenases in SLE patients, were published subsequently.[134] More recently it has become clear that NSAIDs (except possibly naproxen) can predispose to acute myocardial infarction in individuals with coronary heart disease,[322] which is an additional reason for caution in lupus patients.

Conclusions. Based on one systematic review of the evidence from case series and case reports, the overall LOE for NSAIDs in non-renal mild lupus is three and GOR is D.

High-SPF UV-A and UV-B Sunblock in SLE

Summary. There is clear evidence that ultraviolet radiation (UV-A and UV-B) can induce various forms of cutaneous lupus.[287] Patients with systemic lupus without cutaneous features have also been found to have an abnormal reaction to UV irradiation.[323]

Evidence. Sunscreens were shown to prevent discoid and subacute cutaneous lupus rashes in a case series[141] and to reduce systemic features such as renal disease, thrombocytopenia and hospitalization in a cohort study.[136] Three open-label controlled trials,[137–139] a retrospective case series[140] and a double-blind, controlled trial[135] have shown that sunscreens that block UV-A and UV-B can reduce UV radiation–induced lesions of cutaneous lupus.

Conclusions. Lupus patients should be advised about avoidance of sun and other sources of UV irradiation, and about the use of sunscreens (UV-A protection five stars and UV-B protection from SPF factors 30 to 50 products, which can be prescribed on the NHS) and protective clothing. Overall, the LOE is 2++ for sunscreens (one small RCT and six other studies) in lupus patients to prevent cutaneous lesions, and the GOR is B.