Alle patiënten met SLE worden met hydroxychloroquine (HCQ) behandeld, tenzij er een zwaarwegende contra-indicatie is. Bij ziekteactiviteit is er meestal een indicatie voor het starten van een aanvullende behandeling. Vaak worden glucocorticoïden (GC’s) ingezet vanwege hun snelle en effectieve werking. In verband met het ongunstige bijwerkingenprofiel van GC’s worden deze het liefst kortdurend ingezet en snel weer afgebouwd. Indien het niet mogelijk is om de GC’s voldoende af te bouwen, of er sprake is van ernstige, recidiverende of residuale ziekteactiviteit dan dienen ook andere immunosuppressieve medicamenten gestart te worden. De keuze voor een specifiek medicament wordt bepaald door de mate van de ziekteactiviteit, specifieke orgaanmanifestaties, bijwerkingen en comorbiditeit, eventuele zwangerschapswens en voorkeur van de patiënt. Zo wordt er gebruik gemaakt van zonnebrandcrème factor 50 ter preventie, topicale therapie (bijv. glucocorticoïden zalven en tacrolimus zalf) bij patiënten met SLE en met cutane manifestaties als eerstelijns therapie naast HCQ.

In deze module wordt nagegaan welk bewijs beschikbaar is om de arts en patiënt te steunen bij het maken van een keuze bij het behandelen van SLE met cutane manifestaties als tweede- en derdelijns therapie.

 1. csDMARDs

1.1 Oral methotrexate versus placebo

1.2 oral methotrexate versus oral chloroquine

1.3 oral ciclosporin versus oral azathioprine

1.4 oral mycophenolate versus azathioprine or dapsone

1.5 Oral ciclosporin A plus intravenous and oral steroids versus oral steroids alone

2. bDMARDs

2.1 intravenous belimumab (10.0 mg/kg) versus placebo

2.3 intravenous rituximab versus placebo

2.4 Intravenous anifrolumab versus placebo

Description of studies

The Cochrane review by Hannon (2021) investigated the effects of interventions for cutaneous disease in systemic lupus erythematosus (SLE). The Cochrane Skin Information Specialist searched with a predefined strategy in relevant databases (including detailed strategies for CENTRAL, MEDLINE and Embase) and trial registers. The search for relevant studies was performed until September, 8^th 2020. Randomised controlled trials (RCTs) of interventions for cutaneous disease in SLE (according ACR criteria) compared with placebo, another intervention, no treatment, or different doses of the same intervention were eligible for inclusion. Studies including patients without a diagnosis of SLE were excluded. The standard methodological procedures expected by Cochrane were used for data collection and analysis. The primary outcomes were complete and partial clinical response. Clinical response was defined as the percentage of participants with SLE with complete resolution of cutaneous disease based on the Gilliam 1981 classification (lupus-specificor lupus-non-specific) (i.e., absence of (cutaneous) rash). Secondary outcomes included reduction (or change) in number of clinical flares in cutaneous disease, and (severe or minor) adverse events. Severe events were defined as events leading to discontinuation or withdrawal from the study, or resulting in significant morbidity or mortality. Minor events were defined as bothersome to participants but not leading to withdrawal from the study. If the event was further specified, this was described in the result section per comparison.

The GRADE method was used to assess the quality of evidence. In total 61 RCTs were included which reported 43 different interventions, and 35 comparisons. A number of 9 comparisons were of interest for the current research question, as medical treatment with csDMARDs or bDMARDs was administrated. In the comparisons of interest, 12 RCTs were included. These trials were performed in Brazil (n=1), Bangladesh (n=1), Italy (n=1), Malaysia (n=1), North America (i.e., USA/ Canada, n=3), United Kingdom (n=1) and several countries over the world (n=3). Regarding risk of bias, this was highest for the domain of reporting bias, followed by performance/detection bias. Detailed information is provided in the risk of bias table (Hannon, 2021). The trials predominantly included women from outpatient clinics; the mean age range of the participants was 20 to 40 years. Most often treatment duration was 12 months. Due to the limited number of studies per comparison, it was not possible to perform meta-analysis.

Results

Results will be described according to the comparisons of interest in the Cochrane Review of Hannon (2021). Therefore, the summary of literature includes the following comparisons;

1. csDMARDs

1.1 oral methotrexate versus placebo,

1.2 oral methotrexate versus oral chloroquine,

1.3 oral ciclosporin versus oral azathioprine,

1.4 oral mycophenolate versus azathioprine or dapsone,

1.5 oral ciclosporin A plus intravenous and oral steroids versus oral steroids alone.

2. bDMARDs

2.1 intravenous belimumab (1.0, 4.0, 10.0 mg/kg) versus placebo,

2.2 intravenous belimumab versus intravenous belimumab dose comparison,

2.3 intravenous rituximab versus placebo,

2.4 intravenous anifrolumab versus placebo.

1. csDMARDs

1.1 Oral methotrexate versus placebo

absence of malar or discoid rash

One study reported data for the primary outcome of complete clinical response (Carneiro 1999). More participants (17/20) in the methotrexate group than in the placebo group (5/21) had complete clinical resolution of malar or discoid rash at the time of outcome assessment. The risk ratio (RR) was 3.57, 95% confidence interval (CI) 1.63 to 7.84, in favour of the methotrexate group. Importantly, patients used prednisone in both groups.

severe clinical flare

One study reported data for the primary outcome ‘clinical flare’ (Fortin, 2008). More participants (10/45) in the placebo group than in the methotrexate group (7/41) experienced a severe clinical flare requiring discontinuation from the study. This results in a RR of 0.77 (95% CI 0.32 to 1.83).

adverse events – severe

Severe events were defined as events leading to discontinuation or withdrawal from the study, or resulting in significant morbidity or mortality (Hannon, 2021).

One trial reported this outcome on short-term (i.e., 6 months) (Carneiro 1999). In the placebo group, 2/21 participants had lupus flares resulting in withdrawal from the study. The methotrexate group reported 2/20 adverse events (one case of pulmonary tuberculosis, one case of urticaria and dyspepsia). This resulted in an RR of 1.05 (95%CI 0.16 to 6.76).

One trial reported this outcome on longer-term (i.e., 12 months) (Fortin, 2008). Results show that 5/41 in the methotrexate group compared with 0/45 in the placebo group experienced severe adverse effects, resulting in a RR of 12.05 (95%CI 0.69 to 211.36).

adverse events – minor

Minor events were defined as bothersome to participants but not leading to withdrawal from the study (Hannon, 2021).

One trial reported this outcome on short-term (i.e., 6 months) (Carneiro 1999). It was shown that overall, oral methotrexate was more harmful in the total number of minor short-term effects compared with placebo, with more participants in the methotrexate group (11/20) than in the placebo group (3/21) experiencing a minor adverse event, resulting in a RR of 3.85 (95%CI 1.26 to 11.80). For example, regarding gastrointestinal symptoms (mainly dyspepsia and increased hepatic enzyme serum levels), 14 methotrexate patients (70%) presented with side effects compared with only 3 placebo patients (14%).
One trial reported this outcome on longer-term (i.e., 12 months) (Fortin, 2008). Less participants in the methotrexate group (13/41), compared to the placebo group (21/45) experienced an adverse mucocutaneous event (harm). This resulted in an RR of 0.68 (95%CI 0.39 to 1.17).

Level of evidence of the literature

The level of evidence (GRADE method) is determined per comparison and outcome measure and is based on results from RCTs and therefore starts at level “high”. Subsequently, the level of evidence was downgraded if there were relevant shortcomings in one of the several GRADE domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias.

The level of evidence regarding the outcome measures in all comparisons was adapted from the Cochrane review of Hannon (2021). In addition, the level of evidence was downgraded due to imprecision (95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance) in some cases.

The level of evidence regarding the outcome measure complete response: malar/discoid rash was downgraded by two levels due to (1) study design limitations - unclear risk of bias (sequence generation not stated, allocation method not detailed, intent to treat analysis not

stated and could not be confirmed by examination of published data) and (2) imprecision (small study).

The level of evidence regarding the outcome measure clinical flare was downgraded by two levels due to imprecision (small study, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

The level of evidence regarding the outcome measure adverse events (minor and severe) was downgraded by three level due to study design limitations - unclear risk of bias (randomisation method not stated, allocation method not detailed, intent-to-treat analysis not stated and could not be confirmed by examination of published data), and imprecision (2 levels, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

1.2 oral methotrexate versus oral chloroquine

complete clearance of skin rash

One study reported data for the primary outcome ‘complete clearance of skin rash’ (Islam, 2012). In the intention to treat analysis, fewer participants (6/15; 40%) in the methotrexate group compared with the chloroquine group (16/26; 62%) had complete clearance of skin rash (i.e., 6 months). The RR was 0.65, 95%CI 0.33 to 21.30.

In the per protocol analysis, fewer participants (6/13; 46%) in the methotrexate group compared with the chloroquine group (16/24; 67%) had complete clearance of skin rash (i.e., 6 months). The RR was 0.69, 95%CI 0.36 to 1.33.

complete clearance of skin rash during 6-month study (subset of patients with skin findings at baseline)

One study reported data for the primary outcome ‘complete clearance of skin rash’ (Islam, 2012). Results show that 6/6 (100%) participants in the methotrexate group compared with 16/19 (84%) in the chloroquine group had complete clearance of skin rash (subset of patients with skin findings at baseline). The RR was 1.13, 95%CI 0.84 to 1.50.

adverse events – severe

One trial reported this outcome on short-term (i.e., 6 months) (Islam, 2012). In total, 2/15 (13%) participants in the methotrexate group compared with 2/26 (8%) in the chloroquine group had severe adverse events. The RR was 1.73, 95%CI 0.27 to 11.07.

adverse events – minor

One trial reported this outcome on short-term (i.e., 6 months) (Islam, 2012). In total, 9/15 (60%) participants in the methotrexate group compared with 4/26 (15%) in the chloroquine group had severe adverse events. The RR was 3.90, 95%CI 1.45 to 10.51.

Level of evidence of the literature

The level of evidence regarding the outcome measure complete clearance of skin rash during 6-month study (all patients or subset of patients with skin findings at baseline) was downgraded by four levels due to risk of bias (1 level; high risk due to open-label study (no blinding), unclear allocation concealment (1 level; withdrawals occurring without explanation, per protocol analysis), and imprecision (2 levels, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

The level of evidence regarding the outcome measure adverse events (severe and minor) was downgraded by three levels due to risk of bias (1 level; high risk due to open-label study (no blinding), unclear allocation concealment (1 level; withdrawals occurring without explanation, per protocol analysis), and imprecision (1 level, 95%CI of mean difference includes not effect (RR=1), or crosses the threshold of clinical relevance).

1.3 oral ciclosporin versus oral azathioprine

complete clearance of malar rash

One study reported data for the primary outcome of complete clinical response (Griffiths, 2010). Fewer participants (14/47; 30%) in the ciclosporin group compared with the azathioprine group (15/42; 36%) had complete clearance of malar rash at the time of outcome assessment (i.e., 12 months). The RR was 0.83, 95%CI 0.46 to 1.52.

complete clearance of oral ulcers

One study reported data for the primary outcome ‘complete clearance of oral ulcers’ (Griffiths, 2010). More participants (23/47; 49%) in the ciclosporin group compared with the azathioprine group (14/42; 33%) had complete clearance of oral ulcers at the time of outcome assessment (i.e., 12 months). The RR was 1.47, 95%CI 0.87 to 2.46.

adverse events – all severe events combined

One trial reported this outcome on longer-term (i.e., 12 months) (Griffiths, 2010). In the ciclosporin group 14/47 (30%) participant had severe adverse events, compared 14/42 (33%) in the azathioprine group. The RR was 0.89, 95%CI 0.48 to 1.65.

adverse events; subset of severe events due to lack of effect of medication only

One trial reported this outcome on longer-term (i.e., 12 months) (Griffiths, 2010). In total, 5/47 (11%) participants in the ciclosporin group compared with 2/42 (5%) in the azathioprine group had severe adverse events due to lack of effect of medication only (i.e., discontinued treatment due to lack of effect). The RR was 2.23, 95%CI 0.46 to 10.91.

adverse events – minor

One trial reported this outcome on longer-term (i.e., 12 months) (Griffiths, 2010). Data was categorised into six subgroups.

Regarding the AE ‘gum hypertrophy’, results show that 3/47 (6%) participants in the ciclosporin group compared with 0/42 in the azathioprine group had the minor adverse event of gum hypertrophy. The RR was 6.27 (95%CI 0.33 to 117.96).

Regarding the AE ‘alopecia’, results show that 6/47 (13%) participants in the ciclosporin group compared with 7/42 (17%) in the azathioprine group had alopecia. The RR was 0.77 (95%CI 0.28 to 2.10).

Regarding the AE ‘herpes zoster’, results show that 0/47 participants in the ciclosporin group compared with 2/42 (5%) in the azathioprine group had the minor adverse event of zoster infection. The RR was 0.18 (95%CI 0.01 to 3.63).

Regarding the AE ‘non-lupus rash’, results show that 9/47 (19%) participants in the ciclosporin group compared with 12/42 (29%) in the azathioprine group had the minor adverse event of a non-lupus rash. The RR was 0.67 (95%CI 0.31 to 1.43).

Regarding the AE ‘acne’, results show that none of the participants in the ciclosporin group compared with 3/42 (7%) in the azathioprine group had the minor adverse event of acne. The RR was 0.13 (95%CI 0.01 to 2.41).

Regarding the AE ‘hirsutism’, results show that 11/47 (23%) participants in the ciclosporin group compared with 1/42 (2%) in the azathioprine group had the minor adverse event of hirsutism. The RR was 9.83 (95%CI 1.32 to 72.95).

Level of evidence of the literature

The level of evidence regarding the outcome measure complete clearance of malar rash was downgraded by three levels due to risk of bias (1 level; very serious study design limitations from unblinded study), and imprecision (2 levels, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

The level of evidence regarding the outcome measure complete clearance of oral ulcers was downgraded by two levels due to risk of bias (1 level; very serious study design limitations from unblinded study), and imprecision (1 levels, 95%CI of mean difference includes not effect (RR=1)).

The level of evidence regarding the outcome measure adverse events (all severe events combined; subset of severe events due to lack of effect of medication only; minor) was downgraded by three levels due to risk of bias (1 level; very serious study design limitations from unblinded study), and imprecision (2 levels, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

1.4 oral mycophenolate versus azathioprine or dapsone

Importantly study medication was added to standard care (e.g., prednisone, hydroxychloroquine).

partial improvement

One study reported data for the primary outcome ‘partial improvement’ at 4 months (Yahya, 2013). In total, 2 of 3 participants in the mycophenolate group compared with none of 3 in the azathioprine or dapsone group had partial improvement. The RR was 5.0, 95%CI 0.34 to 74.52.

adverse events - severe

One trial reported this outcome on short-term (i.e., 4 months) (Yahya, 2013). None of the 8 participants in the mycophenolate group compared with 2 of 6 patients in the azathioprine or dapsone group had severe adverse events. The RR was 0.16, 95%CI 0.01 to 2.75.

adverse events - minor

One trial reported this outcome on short-term (i.e., 4 months) (Yahya, 2013). In all, 1 of 8

participants in the mycophenolate group compared with none of 6 participants in the azathioprine or dapsone group had minor adverse events. The RR was 2.33, 95%CI 0.11 to 48.99.

Level of evidence of the literature

The level of evidence regarding the outcome measure partial improvement was downgraded by three levels due to risk of bias (1 level; some high-risk domains) and for imprecision (2 levels; one small study, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

The level of evidence regarding the outcome measure adverse events was downgraded by three levels due to risk of bias (1 level; some high-risk domains) and for imprecision (2 levels; one small study, 95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

1.5 Oral ciclosporin A plus intravenous and oral steroids versus oral steroids alone

complete resolution of erythematous manifestations

One study reported data for the primary outcome ‘complete resolution of erythematous manifestations’ over 3 and 12 months (Dammacco, 2000).
In all, 1 of 10 participants in the oral ciclosporin A plus intravenous and oral steroids group compared with none of 8 in the oral steroids alone group had complete resolution of erythematous manifestations over the short term (three months). The RR was 2.45, 95%CI 0.11 to 53.25.

Over the long term, 7 of 10 participants in the oral ciclosporin A plus intravenous and oral steroids group compared with none of 8 in the oral steroids alone group had complete resolution of erythematous manifestations. The RR was 12.27, 95%CI 0.81 to 187.01.

adverse events - severe

One trial reported this outcome over long term (i.e., 12 months) (Dammacco, 2000). In total, 2 of 10 participants in the oral ciclosporin A plus intravenous and oral steroids group compared with 4 of 8 participants in the oral steroids alone group had severe long-term adverse events. The RR was 0.40, 95%CI 0.10 to 1.66.

adverse events – minor

One trial reported this outcome over long term (i.e., 12 months) for mucocutaneous alterations only (hypertrichosis, striae rubrae) and combined mucocutaneous and

other organ systems (Dammacco, 2000).

For the minor AE ‘mucocutaneous alterations only (hypertrichosis, striae rubrae)’, none of 10 participants in the oral ciclosporin A plus intravenous and oral steroids group compared with 4 of 8 in the oral steroids alone group had minor long-term adverse events of mucocutaneous alterations only (hypertrichosis, striae rubrae), favouring oral ciclosporin A plus intravenous and oral steroids. The RR was 0.09, 95%CI 0.01 to 1.47.

For the minor AE ‘combined mucocutaneous and other organ systems’, 6 of 10 participants

in the oral ciclosporin A plus intravenous and oral steroids group compared with 5 of 8 in the oral steroids alone group had minor long-term adverse events of combined mucocutaneous and other organ systems, favouring ciclosporin A plus intravenous and oral steroids due to fewer minor side effects. The RR was 0.96, 95%CI 0.46 to 2.01.

Level of evidence of the literature

The level of evidence regarding the outcome measure complete resolution of erythematous manifestations was downgraded by three levels due to risk of bias (1 level; high risk domains due to the open study (no blinding)) and for imprecision (2 levels; one small study,95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance.

The level of evidence regarding the outcome measure adverse events (severe and minor) was downgraded by three levels due to risk of bias (1 level; high risk domains due to the open study (no blinding)) and for imprecision (2 levels; one small study,95%CI of mean difference includes not effect (RR=1), and crosses both thresholds of clinical relevance).

2. bDMARDs

2.1 intravenous belimumab (10.0 mg/kg) versus placebo

Three studies compared intravenous belimumab with placebo (Furie, 2011; Navarra, 2011; Wallace, 2009), and one study compared subcutaneous belimumab with placebo (Stohl, 2017). All patients continued standard care, and belimumab was added in the intervention group. The definition of standard care was ‘continue new antimalarial or current medication’ in the study of Furie (2011), and ‘with changes not permitted after 16 weeks for immunosuppressant

agents and 24 weeks for antimalarials’ in the study of Navarra (2011). Wallace (2009) and Stohl (2017) did not specify the definition.

skin-specific measures of SLE disease activity (SELENA/SLEDAI, BILAG, and SRI)

 More participants in the belimumab 10 mg/kg group had "SELENA/SLEDAI scores reduced by 4 points" in the mucocutaneous domain compared with the placebo group (OR 1.71, 95% CI 1.21 to 2.41; P<0.01). More participants in the belimumab 10 mg/kg dose group had "no new BILAG A" in the mucocutaneous domain or "no more than 1 new BILAG B flare" in the mucocutaneous domain than did those in the placebo group (OR 1.62, 95% CI 1.09 to 2.42; P = 0.02

Stohl (2017) reported on subcutaneous belimumab, noting that "all components of the SRI4 (Systemic Lupus Erythematosus Responder Index (SRI4)) showed statistical significance at week 52" and that "the immunologic, musculoskeletal, mucocutaneous, and vascular SELENA–SLEDAI organ systems” improved significantly more with belimumab.

adverse events; severe, at belimumab 10mg/kg dose

Two trials reported this outcome over 12 months (Navarrra, 2011; Wallace, 2009). Severe adverse events occurred in 33/401 (8%) participants in the belimumab group, compared with 41/400 (10%) in the placebo group. This resulted in a RR of 0.81, 95%CI 0.53 to 1.24.

One trial reported this outcome over 18 months (Furie, 2011). Severe adverse events occurred in 23/273 (8%) participants in the belimumab group, compared with 23/275 (8%) in the placebo group. This resulted in a RR of 1.01, 95%CI 0.58 to 1.75.

By comparing the different timepoints, severe adverse events occurred in 56/674 (8%) participants in the belimumab group compared to 64/675 (9%) in the placebo group. This resulted in an overall RR of 0.88 (95%CI 0.63 to 1.23).

adverse events; minor- skin and subcutaneous skin combined data for doses belimumab

One trial reported this outcome over 12 months (Wallace, 2009). In total, with 192 out of 336 participants in the belimumab groups, compared with only 57 out of 113 participants (50%) in the placebo group experienced minor "skin and subcutaneous skin" adverse events (i.e., skin and subcutaneous tissue disorders) over 12 months. Specifically, for the belimumab subgroups, minor adverse events were experienced in 72 out of 114 participants (63%) in the 1 mg/ kg group, in 65 out of 111 participants (59%) in the 4 mg/kg group, and in 55 out of 111 participants (50%) in the 10.0 mg/kg group, respectively. This resulted in an RR of 1.13, 95%CI 0.92 to 1.39, suggesting more minor adverse events in the belimumab groups compared to placebo.

Stohl (2017) reported that minor adverse events occurred in 389/556 (68%) participants in the belimumab (SC) group, compared with 182/280 (65%) in the placebo group.

Level of evidence of the literature

The level of evidence regarding the outcome measure skin-specific measures of SLE disease activity, adverse events (severe) was downgraded by two levels due to risk of bias (1 level; insufficient details about randomisation), and imprecision (1 level; 95%CI of mean difference includes not effect (RR=1), or crosses both thresholds of clinical relevance.

The level of evidence regarding the outcome measure adverse events (minor - skin and subcutaneous skin) was downgraded by two levels due to risk of bias (1 level; insufficient details about randomisation), and imprecision (1 level; 95%CI of mean difference includes not effect (RR=1), or crosses both thresholds of clinical relevance.

2.3 intravenous rituximab versus placebo

skin-specific measures of SLE disease activity (BILAG)

One trial reported the outcome measure including ‘BILAG B and BILAG A’ scores for the mucocutaneous domain (Merrill, 2010). At baseline in the rituximab group, 56.2% had BILAG B and 16% had BILAG A. At baseline in the placebo group, 58.0% had BILAG B and 13.6% had BILAG A. Subgroup analysis of the primary endpoint (BILAG A and BILAG B) was performed for the preplanned subgroup "mucocutaneous system involvement", and study investigators

reported no differences during the 12-month study. Standard care (i.e., prednisone plus a baseline immunosuppressant) was continued in both groups.

adverse events; severe – death

One trial reported this outcome over long term (i.e., 12 months) (Merrill, 2010). A total of 4 of 169 (2%) participants in the rituximab group compared with 1 of 88 (1%) participants in the placebo group died. The RR was 2.08, 95%CI 0.24 to 18.85.

adverse events; severe - event not including death

One trial reported this outcome over long term (i.e., 12 months) (Merrill, 2010). In all, 19 of 169 (5%) participants in the rituximab group compared with 13 of 88 (15%) in the placebo group had severe adverse events (not including death). The RR was 0.76, 95%CI 0.39 to 1.47.

Level of evidence of the literature

The level of evidence regarding the outcome measure skin-specific measures of SLE disease activity, adverse events (severe - death) was downgraded by one level due to risk of bias (1 level; insufficient details about methods of randomisation and allocation concealment).

The level of evidence regarding the outcome measure adverse events (severe – death; severe - not including death) was downgraded by three level due to risk of bias (1 level; insufficient details about methods of randomisation and allocation concealment), and imprecision (2 levels; 95%CI of mean difference includes not effect (RR=1), or crosses both thresholds of clinical relevance).

2.4 Intravenous anifrolumab versus placebo

Data for this comparison was limited since exact details about number of patients with active skin involvement were not given in the trials. Only baseline Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI) scores, and the number of patients with improved CLASI scores after treatment with anifrolumab were reported. No information was provided for adverse events in the specific groups of interest. Information about adverse events in all patients are reported in ‘module Ziekteactiviteit’. Importantly, in this review patients were administrated to placebo or 300 mg anifrolumab via intravenous infusion every 4 weeks, and patients in both groups received standard care.

skin-specific measures of SLE disease activity: CLASI activity score

Three trials reported ≥50% improvement in CLASI in patients who had a score of ≥10 at baseline (Furie 2017;2019; Morand, 2020).

One of them reported this outcome after 12 weeks, resulting in a RR of 2.43 (95%CI 1.20 to 4.93), as 24/180 (49%) achieved this outcome in the anifrolumab group, compared with 10/182 (25%) in the control group (Morand, 2020).

Another study reported this outcome after 24 weeks, resulting in a RR of 2.05 (95%CI 1.07 to 3.90), as 17/27 (63%) achieved this outcome in the anifrolumab group, compared with 8/26 (31%) in the control group (Furie, 2017).

The last study reported this outcome after 52 weeks, resulting in a RR of 1.60 (95%CI 0.93 to 2.75), as 24/58 (42%) achieved this outcome in the anifrolumab group, compared with 14/54 (25%) in the control group (Furie, 2019).

Level of evidence of the literature

The level of evidence regarding the outcome measure skin-specific measures of SLE disease activity (≥50% improvement in CLASI) was downgraded by one level due to risk of bias (domains unclear risk or high risk).

A systematic review of the literature was performed to answer the following question:

What are the benefits and harms of treatment with csDMARDs or bDMARDs compared to placebo or control treatment in adults with Systemic Lupus Erythematosus on absence of malar, subacute cutaneous lupus erythematosus, discoid rash, or adverse events?

P:           adults with Systemic Lupus Erythematosus (SLE)

I:            medical treatment with csDMARDs or bDMARDs

C:           placebo or control treatment

O:          absence of malar, subacute cutaneous lupus erythematosus, discoid rash, adverse events.

Relevant outcome measures

The guideline development group considered absence or significant improvement of malar, subacute cutaneous lupus erythematosus, discoid rash as a critical outcome measure for decision making; and adverse events as an important outcome measure for decision making.

A priori, the working group did not define the outcome measures listed above but used the definitions used in the studies.

A difference of 25% in the relative risk for dichotomous outcomes (i.e., RR 0.80-1.25) and 0.5 standard deviation (reported as SMD) for continuous outcomes was taken as a minimal clinically important difference.

Search and select (Methods)

The databases Medline (via OVID) and Embase (via Embase.com) were searched with relevant search terms until November 28^th 2020 for systematic reviews and randomized controlled trials (RCTs). The detailed search strategy is depicted under the tab Methods. After removing duplicates the combined systematic literature search resulted in 10067 hits (2020 systematic review, 8047 RCT). Considering the total amount of hits, it was decided to first only use the systematic review selection. Studies were selected based on the following criteria;

- adults with Systemic Lupus Erythematosus (SLE),

- medical treatment with csDMARDs (i.e., cyclophosphamide, azathioprine, methotrexate, leflunomide, mycophenolate mofetil (MMF), cyclosporin A, and tacrolimus) or bDMARDs (i.e., anifrolumab, belimumab (BEL), rituximab (RTX)) compared to placebo of control treatment,

- outcome according to the PICO (i.e., absence/improvement of malar, subacute cutaneous lupus erythematosus, discoid rash, adverse events) was studied.

In total 145 systematic reviews were initially selected based on title and abstract screening. After a full text screening 121 studies were excluded and 24 eligible systematic reviews for the present guideline were included. After reading 24 articles full text, nine systemic reviews described relevant outcomes for the current clinical question. Eight of them were excluded (see the table with reasons for exclusion under the tab Methods). The most recent and complete systemic review was selected for the literature analysis.

Results

One study was included in the analysis of the literature. Important study characteristics and results are summarized in the evidence tables. The assessment of the risk of bias is summarized in the risk of bias tables.