Combining mepacrine with hydroxychloroquine-based therapy in active systemic lupus erythematosus: an observational study of 106 patients from the Lupus-Cruces cohort Authors: Beatriz Marín-García, Luis Dueña-Bartolomé, Oihane Montero, Guillermo Ruiz-Irastorza Published: 2025 ABSTRACT AbstractObjectiveTo evaluate the efficacy of mepacrine (MC) as an add-on therapy in patients with SLE unresponsive to hydroxychloroquine (HCQ)-containing regimens at 6 months after MC introduction.MethodsObservational study using routine clinical care data of patients from the Lupus-Cruces cohort. All of them received therapy with HCQ and prednisone at baseline. Two groups of initial therapy were compared: single therapy (ST; HCQ + prednisone) and multiple therapy (MT; additional immunosuppressives or biologics). Achieving the definition of remission in SLE (DORIS) at 6 months was the main outcome. Prednisone tapering and MC side effects and discontinuation were also analysed. A logistic regression was performed in search of clinical predictors of response.Results106 different episodes were included (ST=56, MT=50). The mean SLE Disease Activity Index (SLEDAI) at baseline was 6.7, with a mean prednisone dose of 6 mg/day. DORIS remission at 6 months was 71% for the complete cohort (ST 79% vs MT 62%, p=0.06). SLEDAI reduction at 6 months was similar in both groups (mean 4.6 points in the ST group vs 5 in the MT group, p=0.5). The reduction at 6 months was also similar (mean 1.75 mg/day in the ST group vs 1.69 mg/day in the MT group, p=0.9). The most frequent reason for MC discontinuation was improvement (45%). Adverse effects were reported in 17% patients.ConclusionsMC is a useful therapy in mild–moderate active SLE. Using MC as the first drug after the failure of glucocorticoids and HCQ is the best option; however, the addition of MC to a multidrug regimen can also be of help.Keywords: Glucocorticoids; Lupus Erythematosus, Systemic; Arthritis; Antirheumatic Agents WHAT IS ALREADY KNOWN ON THIS TOPIC While treatment with hydroxychloroquine (HCQ) is considered the cornerstone of therapy for SLE, currently the role of mepacrine (MC) has been relegated mainly to cutaneous SLE or to patients with ocular toxicity due to HCQ. While treatment with hydroxychloroquine (HCQ) is considered the cornerstone of therapy for SLE, currently the role of mepacrine (MC) has been relegated mainly to cutaneous SLE or to patients with ocular toxicity due to HCQ. Combined antimalarial therapy with HCQ and MC has shown higher efficacy to treat active lupus, not only skin disease, in small series. Combined antimalarial therapy with HCQ and MC has shown higher efficacy to treat active lupus, not only skin disease, in small series. WHAT THIS STUDY ADDS Adding MC to an HCQ-based treatment in patients with articular, cutaneous and/or serosal lupus activity resulted in a 71% definition of remission in SLE (DORIS) remission rate at 6 months. Adding MC to an HCQ-based treatment in patients with articular, cutaneous and/or serosal lupus activity resulted in a 71% definition of remission in SLE (DORIS) remission rate at 6 months. Patients with baseline therapy with HCQ + low dose prednisone were more likely to respond than those receiving additional immunosuppressive ± biological therapy. However, more than 60% of the latter achieved DORIS remission without the need for additional therapy. Patients with baseline therapy with HCQ + low dose prednisone were more likely to respond than those receiving additional immunosuppressive ± biological therapy. However, more than 60% of the latter achieved DORIS remission without the need for additional therapy. Glucocorticoid-sparing effects were also seen after adding MC. Glucocorticoid-sparing effects were also seen after adding MC. Side effects were few and did not result in drug withdrawal in most affected patients. Side effects were few and did not result in drug withdrawal in most affected patients. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY The early addition of MC to patients with active articular, cutaneous and/or serosal SLE is an effective and safe alternative to more toxic and/or costly treatments. The early addition of MC to patients with active articular, cutaneous and/or serosal SLE is an effective and safe alternative to more toxic and/or costly treatments. The additional use of MC can help keep long-term doses of HCQ within the safe threshold for ocular toxicity. The additional use of MC can help keep long-term doses of HCQ within the safe threshold for ocular toxicity. These results and further research could impulse the future use and availability of MC as a safe and affordable alternative to other second-line drugs in SLE. These results and further research could impulse the future use and availability of MC as a safe and affordable alternative to other second-line drugs in SLE. Introduction Hydroxychloroquine (HCQ) is the cornerstone of therapy for SLE, exerting a wide range of beneficial effects, from modulating the autoimmune response without immunosuppression to thromboprophylaxis.1 The results of prolonged antimalarial therapy in SLE include preventing damage accrual and, eventually, increasing survival.2 Therefore, it is now common consensus that all patients with SLE should receive HCQ unless contraindicated.3 However, despite being an essential piece in the management of SLE, HCQ (± low dose prednisone) is often not enough to keep SLE activity under control. In such a scenario, additional therapy with immunosuppressives (ISs) ± biological drugs is recommended in recent guidelines,3 in which the role of mepacrine (MC) has been limited to patients with cutaneous lupus. In 2018, we first published a case series of 46 patients showing the efficacy and safety of MC in patients with SLE presenting with refractory joint or cutaneous manifestations.4 Encouraged by the good results, we have increasingly used MC as an add-on therapy in patients with maintained articular, cutaneous or serosal activity despite optimal background therapy with HCQ plus low-dose prednisone, thus expanding both the number of patients treated with MC and the follow-up time. In this study, we aim to update the clinical utility of adding MC to patients with active SLE despite conventional HCQ + prednisone-based therapy. Objectives The main objective was to evaluate the efficacy of MC as an add-on therapy in patients with SLE with cutaneous, articular and/or serosal manifestations unresponsive to HCQ-containing regimens. As secondary objectives, we aimed to: Identify the optimal treatment stage to add MC, either as the first drug after glucocorticoids (GCs) and HCQ or to a multidrug scheme containing GCs, HCQ, IS ± biological drugs. Identify the optimal treatment stage to add MC, either as the first drug after glucocorticoids (GCs) and HCQ or to a multidrug scheme containing GCs, HCQ, IS ± biological drugs. Evaluate the safety of MC and the reasons for withdrawal. Evaluate the safety of MC and the reasons for withdrawal. Methods Study design and patients We conducted an observational study using routine clinical care data. Within the Lupus-Cruces database, we identified 103 patients receiving MC as an add-on therapy for insufficient control of lupus activity while on therapy with at least HCQ and prednisone. As three patients received two different courses of MC, we finally analysed 106 different episodes of MC use. All the patients included in this study fulfilled the American College of Rheumatology/Systemic Lupus International Collaborating Clinics criteria for the classification of SLE.5 6 Baseline variables Variables reflecting the demographic, clinical and immunological status of the patients at the time of MC initiation included: age, gender, ethnicity, disease duration, smoking status, autoantibody profile, complement levels, main cumulative clinical manifestations of SLE, main indication for MC therapy (joint, skin or serosal activity) and degree of disease activity as measured by SLE Disease Activity Index 2000 (SLEDAI-2K) score,7 categorised as mild (SLEDAI <6 points), moderate (SLEDAI 6–12) or severe activity (SLEDAI >12).3 Also, data on therapy at the time of MC introduction were analysed, including MC starting dose, prednisone and HCQ doses and additional therapy with IS: methotrexate (MTX), azathioprine (AZA), mycophenolate (MMF), cyclophosphamide or tacrolimus; and with biological drugs: rituximab or belimumab (BEL). Groups of initial therapy According to the selection criteria, all patients included in the study were receiving baseline therapy with HCQ and prednisone. However, as other additional therapies could have been used before taking the decision of adding MC, two different therapeutic groups were identified: those receiving at baseline HCQ and prednisone only (named as the single therapy group, or ST, n=56) and those treated with other IS and/or biological drugs in addition to HCQ and prednisone (the multiple therapy group, or MT, n=50). These groups were used to investigate the best stage to introduce MC within the process of therapy escalation. Outcomes The main efficacy outcome was achieving definition of remission in SLE (DORIS) remission within 6 months of MC introduction.8 Patients who stopped MC therapy for any reason before 6 months and had not yet achieved remission at that time were classified as ‘DORIS non-responders’. The use of additional therapy after starting MC, except for additional methylprednisolone pulses (MP), was also a classification criterion for ‘DORIS non-responder’. In addition, two secondary efficacy outcomes were analysed: Clinical response to MC as measured by the responsible clinicians, as follows:Complete remission: complete resolution of the symptoms that motivated MC introduction without the use of additional therapy after starting MC, except for MP.Partial remission (PR): improvement of the symptoms, with no complete resolution, and/or requirement of additional therapy after MC initiation. Clinical response to MC as measured by the responsible clinicians, as follows: Complete remission: complete resolution of the symptoms that motivated MC introduction without the use of additional therapy after starting MC, except for MP. Complete remission: complete resolution of the symptoms that motivated MC introduction without the use of additional therapy after starting MC, except for MP. Partial remission (PR): improvement of the symptoms, with no complete resolution, and/or requirement of additional therapy after MC initiation. Partial remission (PR): improvement of the symptoms, with no complete resolution, and/or requirement of additional therapy after MC initiation. No response: failure of achieving any clinical improvement after MC initiation. No response: failure of achieving any clinical improvement after MC initiation. Prednisone dose reduction after MC initiation. Prednisone dose reduction after MC initiation. Unlike our previous work,4 we decided not to include the Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI) as an additional outcome, for a number of reasons: first, cutaneous lupus is the subset in which the effects of MC have been already better defined; second, CLASI is not routinely performed in our clinics; third, patients with cutaneous disease as the main indication for MC were a minority in this cohort; and lastly, we aimed to focus on a global response concept in patients with systemic lupus.8 Finally, the occurrence of MC-related side effects, MC discontinuation and reasons for discontinuation (improvement, lack of effect, toxicity, pregnancy, patient’s choice, other) were also recorded. Statistical analysis Descriptive data were generated, using percentages, means and SD, for the whole cohort and for the two subgroups constructed according to the baseline therapy scheme (ST and MT). Data were compared by χ² or Student’s t-test, as appropriate. For the main outcome, the rate of DORIS remission at 6 months after MC introduction was calculated for the whole cohort. In addition, the effect of MC in reducing SLEDAI scores and prednisone doses between baseline and 6 months was assessed by using paired Student’s t-test. In order to identify possible predictors of response to MC, baseline variables, including the two therapeutic subgroups (ST and MT), were compared by χ² or Student’s t-test according to the achievement of DORIS remission at 6 months. All variables with a p value ≤0.2 in the univariate analysis were entered into a logistic regression model with DORIS remission at 6 months as the dependent variable, with sequential withdrawal of non-significant variables in a backward stepwise way. The statistical analysis was done using Stata/MP V.18 for Mac (StataCorp LP, Texas, USA). Patient and public involvement Patients or the public were not involved in the design, conduct, reporting or dissemination plans of our research. Results Baseline variables We analysed 106 episodes of MC therapy in 103 patients, 3 patients receiving two different courses of MC each (table 1). Most patients were women (89.6%) and white (85.8%), with a mean age of 40.7 years. The median disease duration was 5 years in both groups. The most frequently found antibodies were anti-double-stranded DNA (dsDNA) (62.3%) followed by antiphospholipid antibodies (60.4%) and anti-Ro (49.1%). Table 1Baseline clinical, immunological and therapeutic characteristicsTotal (N=106)ST (n=56)MT (n=50)P valueFemale95 (89.6%)53 (94.6%)42 (84%)0.07Male11 (10.3%)3 (5.4%)8 (16%)Age (years); mean (SD)40.7 (13.4)44 (14.2)37 (11.8)0.007Disease duration (years); median (IQR)5 (1–11)5 (1–12)5 (2–11)0.9White91 (85.8%)48 (85.7%)43 (86%)0.96Smoking status0.6 No56 (52.8%)27 (48.2%)29 (58%) Active33 (31.1%)19 (33.9%)14 (28%) People who have quit smoking17 (16%)10 (17.9%)7 (14%)Autoantibodies ever Anti-dsDNA66 (62.3%)29 (51.8%)37 (74%)0.01 Anti-Ro52 (49.1%)24 (42.9%)28 (56%)0.17 Anti-La13 (12.3%)8 (14.3%)5 (10%)0.5 Anti-Sm25 (23.6%)11 (19.6%)14 (28%)0.31 Anti-RNP32 (30.2%)13 (23.2%)19 (38%)0.09 aPL64 (60.4%)34 (60.7%)30 (60%)0.94Hypocomplementaemia ever71 (67%)35 (62.5%)36 (72%)0.29Clinical features ever Arthritis80 (75.5%)48 (85.7%)32 (64%)0.009 Skin23 (21.7%)9 (16.1%)14 (28%)0.13 Serositis11 (10.4%)6 (10.7%)5 (10%)0.9 Haematological51 (48.1%)20 (35.7%)31 (62%)0.007 Nephritis20 (18.9%)4 (7.1%)16 (32%)0.001Main indication for MC Arthritis75 (70.8%)43 (76.8%)32 (64%) Skin22 (20.8%)8 (14.3%)14 (28%)0.22 Serositis9 (8.4%)5 (8.9%)4 (8%)Baseline SLEDAI; mean (SD)6.7 (3.4)6.1 (2.9)7.5 (3.9)0.04Baseline SLEDAI Mild (<6)37 (34.9%)21 (37.5%)16 (32%) Moderate (6–12)63 (59.4%)34 (60.7%)29 (58%)0.18 Severe (>12)6 (5.7%)1 (1.8%)5 (10%)Baseline therapiesPrednisone dose (mg/day); mean (SD)6 (2.7)6.1 (2.8)5.9 (2.7)0.77MP pulses58 (55%)36 (64%)22 (44%)0.036Number of MP pulses; mean (SD)*3.01 (0.92)3.05 (0.82)2.95 (1.09)0.691MP cumulative dose (mg); mean (SD)*528.19 (293.16)523.61 (268.39)535.68 (336.34)0.881HCQ dose at baseline 200 mg/day94 (89%)52 (93%)42 (84%)0.15 >200 mg/day12 (11%)4 (7%)8 (16%)MC initial dose 50 mg/day39 (37%)26 (46%)13 (26%)0.029 100 mg/day67 (63%)30 (54%)37 (74%)MC initial dose (mg/day); mean (SD)81.6 (24.2)76.7 (25.1)87 (22.1)0.03IS50 (47.2%) MTX26 (24.5%)0 (0%)26 (52%)<0.001 MMF15 (14.2%)0 (0%)15 (30%)<0.001 AZA13 (12.3%)0 (0%)13 (26%)<0.001 CYC7 (6.6%)0 (0%)7 (14%)0.004 TACR3 (2.8%)0 (0%)3 (6%)0.06Combined IS therapy12 (11.3%)0 (0%)12 (24%)<0.001Biologics6 (5.6%)0 (0%)6 (12%) RTX1 (0.9%)0 (0%)1 (2 %)0.28 BEL5 (4.7%)0 (0%)5 (10%)0.01*Among patients receiving MP pulses.aPL, antiphospholipid antibodies; AZA, azathioprine; BEL, belimumab; CYC, cyclophosphamide; dsDNA, double-stranded DNA; HCQ, hydroxychloroquine; IS, immunosuppressive; MC, mepacrine; MMF, mycophenolate; MP, methylprednisolone; MT, multiple therapy group; MTX, methotrexate; RTX, rituxumab; SLEDAI, SLE Disease Activity Index; ST, single therapy group; TACR, tacrolimus. Baseline clinical, immunological and therapeutic characteristics Among patients receiving MP pulses. aPL, antiphospholipid antibodies; AZA, azathioprine; BEL, belimumab; CYC, cyclophosphamide; dsDNA, double-stranded DNA; HCQ, hydroxychloroquine; IS, immunosuppressive; MC, mepacrine; MMF, mycophenolate; MP, methylprednisolone; MT, multiple therapy group; MTX, methotrexate; RTX, rituxumab; SLEDAI, SLE Disease Activity Index; ST, single therapy group; TACR, tacrolimus. Regarding the cumulative clinical profile of patients, arthritis (75.5%) was the most frequent clinical manifestation followed by immune cytopenias (48.1%) and cutaneous lupus (21.7%). The main indication for adding MC was arthritis (70.8%) followed by skin rash (20.8%) and serositis (8.4%). According to the baseline SLEDAI score, more than half of the patients had moderate activity at baseline (59.4%). According to the inclusion criteria, all patients were under therapy with HCQ 200 mg/day plus prednisone, at a mean dose of 6 mg/day. IS were used in 47.2% patients before MC introduction, 12 patients (11.3%) receiving combined therapy with more than one IS drug. MTX (24.5%) and MMF (14.2%) were the most frequently used IS. Combined therapy including biologics was given to six patients before MC introduction (5.6%), BEL in five of them. Baseline differences between groups of initial therapy Patients in the ST entered the Lupus-Cruces cohort later (50% after 2013, vs 50% after 2010 in the MT groups), reflecting the progressively earlier indication of MC in the therapeutic protocols of our unit over the recent years. The follow-up time was, thus, lower in this group (mean 62 vs 87 months in the MT group, p=0.001). Patients in the ST group were older (mean age 44 years vs 37 years in the MT group), with a higher frequency of women (94.6% vs 84%) and a lower positivity for anti-dsDNA antibodies (51.8% vs 74%). Arthritis was the main indication for MC introduction in both groups, with a non-significant higher frequency of skin disease among patients in the MT group (28% vs 14.3%). The baseline SLEDAI was lower among patients in the ST group (mean 6.1 vs 7.5 in the MT group, p=0.04). However, more than half of the patients in both groups were classified as having moderate activity, with only six patients presenting with severe activity (one in the ST and five in the MT group). The doses of prednisone at baseline were similar between groups (ST 6.1 mg/day vs MT 5.9 mg/day, p=0.77). However, more patients in the ST group received MP at the time of starting MC (64% vs 44%, p=0.036), at the same dose of 125–250 mg/day ×3 days (table 1). Initial MC doses of 100 mg/day were used in 54% patients in the ST group, versus 74% in the MT group (p=0.029), resulting in a lower mean MC starting dose in the ST group (76.7 vs 87 mg/day). Clinical response to MC Within the whole cohort, DORIS remission at 6 months after MC introduction was achieved by 75 patients (71%). Among the 31 patients who did not fulfil DORIS criteria for remission, 14 were classified as PR by clinicians. According to the main indication for MC therapy, patients with articular, cutaneous and serosal manifestations achieved similar rates of DORIS remission (71%, 68% and 78%, respectively, p=0.86). The mean SLEDAI score decreased significantly in the whole group from 6.7 at baseline to 1.9 at 6 months (p<0.001). Likewise, the mean prednisone dose was significantly reduced from 6 to 4.3 mg/day (p<0.001). Clinical response by groups of initial therapy Patients in the ST group were more likely to achieve DORIS remission than those in the MT group (79% vs 62%, respectively, p=0.06). Within the group not achieving DORIS remission, more patients in the MT group were classified as DORIS non-responders for receiving additional therapy after MC initiation or for stopping MC before 6 months (14 patients, 28%, vs 11 patients, 14%, in the ST group, p=0.08). On the other hand, the reduction in the SLEDAI score at 6 months was similar in both groups (mean 4.6 points in the ST group vs 5 in the MT group, p=0.5). The reduction in the dose of prednisone at 6 months was also similar between groups (mean 1.75 mg/day in the ST group vs 1.69 mg/day in the MT group, p=0.9). The mean dose of prednisone at 6 months was <5 mg/day in both groups, with only four patients in the whole cohort (3.7%) receiving doses >5 mg/day (ranging between 6.75 and 10 mg/day). The efficacy outcomes are summarised in table 2. Table 2Efficacy outcomesTotal (N=106)ST (n=56)MT (n=50)P valueDORIS remission75 (71%)44 (79%)31 (62%)0.06SLEDAI at 6 months; mean (SD)1.9 (2.1)1.6 (1.9)2.3 (2.4)0.09SLEDAI reduction at 6 months; mean (SD)4.8 (4.1)4.6 (2.8)5 (4.1)0.5Prednisone dose (mg/day) at 6 months; mean (SD)4.3 (1.4)4.35 (1.3)4.26 (3.8)0.74Prednisone dose reduction at 6 months (mg/day); mean (SD)1.72 (3.1)1.75 (2.9)1.69 (3.24)0.9Need for additional therapy17 (16%)6 (11%)11 (22%)0.11Time of follow-up after starting MC (months); mean (SD)74 (40)62 (34)87 (41)0.001Months on MC therapy; mean (SD)55 (44)45 (37)66 (48.5)0.013Mepacrine dose (mg/day) at the end of follow-up; mean (SD)73 (26)73 (28)74 (25)0.81MC discontinuation47 (44%)25 (44%)22 (44%)0.9Reason for discontinuation0.63 Improvement21/47 (45%)13/25 (52%)8/22 (36%) No response4/47 (8%)1/25 (4%)3/22 (14%) Pregnancy9/47 (19%)4/25 (16%)5/22 (23%) Side effects7/47 (15%)3/25 (12%)4/22 (18%) Other*6/47 (13%)4/25 (16%)2/22 (9%)*Two because of discontinuation in another hospital, two because of personal choice and two due to loss of follow-up.DORIS, definition of remission in SLE; MC, mepacrine; MT, multiple therapy group; SLEDAI, SLE Disease Activity Index; ST, single therapy group. Efficacy outcomes Two because of discontinuation in another hospital, two because of personal choice and two due to loss of follow-up. DORIS, definition of remission in SLE; MC, mepacrine; MT, multiple therapy group; SLEDAI, SLE Disease Activity Index; ST, single therapy group. Predictors of response to MC The univariant analysis for the baseline predictors of DORIS response to MC is shown in table 3. No differences in the rates of DORIS remission were found according to the baseline level of SLE activity, the dose of prednisone, the dose of HCQ, the starting dose of MC or the main indication for MC initiation. Of note, the frequency of active smokers was similar between responders and non-responders (33% vs 26%). On the other hand, non-white patients, those receiving MT, those not receiving MP and those treated with MTX tended to respond less. Table 3Univariate predictors of DORIS remission at 6 monthsDORIS remission (n=75)No DORIS remission (n=31)P valueFemale67 (89%)28 (90%)0.87Age (years); mean (SD)41 (14)38.6 (10)0.30Disease duration (years); median (IQR)5 (1–11)5 (1–12)0.78White67 (89%)24 (77%)0.10People who actively smoke25 (33%)8 (26%)0.44Baseline SLEDAI; mean (SD)6.7 (3.5)7 (3.3)0.70Baseline activity0.68 Mild30 (40%)10 (32.2%) Moderate39 (52%)19 (61.3%) Severe6 (8%)2 (6.5%)Main indication for MC0.86 Arthritis53 (71%)22 (71%) Skin15 (20%)7 (23%) Serositis7 (9%)2 (6%)Initial therapy group0.06 ST44 (59%)12 (39%) MT31 (41%)19 (61%)Prednisone dose at baseline (mg/day); mean (SD)6.14 (2.9)5.8 (2.2)0.5MP at baseline44 (59%)14 (45%)0.2HCQ dose at baseline >200 mg/day9 (12%)3 (10%)0.7MC starting dose (mg/day); mean (SD)81 (24)82 (24)0.85IS at baseline MTX15 (20%)11 (35%)0.09 MMF10 (13%)5 (17%)0.70 AZA9 (12%)4 (13%)0.89 CYC4 (5%)3 (10%)0.41 TACR2 (3%)1 (3%)0.87Belimumab at baseline3 (4%)2 (6.5%)0.62AZA, azathioprine; CYC, cyclophosphamide; DORIS, definition of remission in SLE; HCQ, hydroxychloroquine; IS, immunosuppressive; MC, mepacrine; MMF, mycophenolate; MP, methylprednisolone pulses; MT, multiple therapy group; MTX, methotrexate; SLEDAI, SLE Disease Activity Index; ST, single therapy group; TACR, tacrolimus. Univariate predictors of DORIS remission at 6 months AZA, azathioprine; CYC, cyclophosphamide; DORIS, definition of remission in SLE; HCQ, hydroxychloroquine; IS, immunosuppressive; MC, mepacrine; MMF, mycophenolate; MP, methylprednisolone pulses; MT, multiple therapy group; MTX, methotrexate; SLEDAI, SLE Disease Activity Index; ST, single therapy group; TACR, tacrolimus. In the multivariant logistic regression analysis including these latter four variables in the baseline model, only MT (OR 0.43, 95% CI 0.18 to 1.03, p=0.06) and non-white ethnicity (OR 0.39, 95% CI 0.12 to 1.2, p=0.10) showed an independent trend for a reduced rate of DORIS response at 6 months in the final model. Safety issues and MC discontinuation Patients in the MT group were longer under MC therapy than patients in the ST group (mean 66 months vs 45, respectively, p=0.013), although this was influenced by the longer follow-up time of the MT group (see above). The dose of MC at the end of follow-up was very similar between both groups (72.7 mg/day vs 74 mg/day, p=0.81). There were no differences regarding the frequency of MC discontinuation for any reason (44% in both groups, p=0.9, table 2). The most frequent reason for MC discontinuation was improvement (45%). Adverse effects were seen in 17% patients; however, only seven patients (6.6% of the whole cohort) discontinued MC because of toxicity (table 4). Gastric upset and severe yellowish skin discolouration were the most frequent side effects. Severe side effects such as central anaemia or lichenoid rash (without associated bone marrow toxicity) were only observed in one patient each. No differences between treatment groups were seen. Table 4Side effectsTotalST (n=56)MT (n=50)Side effects18/106 (17%)8/56 (14.3%)10/50 (20%)Side effects causing discontinuation7/106 (6.6%)3/56 (5.3%)4/50 (8%)Type of side effects GI intolerance12/18 (66.7%)5/8 (62.5%)7/10 (70%) Severe yellowish skin discolouration3/18 (16.7%)1/8 (12.5%)2/10 (20%) Transaminase elevation1/18 (5.6%)0 (0%)1/10 (10%) Neurological (headache)1/18 (5.6%)1/8 (12.5%)0 (0%) Haematological (central anaemia)1/18 (5.6%)1/8 (12.5%)0 (0%)GI, gastrointestinal; MT, multiple therapy group; ST, single therapy group. Side effects GI, gastrointestinal; MT, multiple therapy group; ST, single therapy group. Discussion In this study, we have confirmed that MC is a useful add-on therapy in patients with refractory articular, cutaneous and/or serosal SLE. Combined therapy with MC caused few and mild side effects, rarely leading to drug discontinuation and allowed the reduction of prednisone doses to ≤5 mg/day in all responding patients. MC (also known as atabrine or quinacrine) was produced in Germany during the 1920s and widely used for malarial prophylaxis in the Pacific theatre during World War II.9 10 After the empirical observation of British soldiers with inflammatory conditions improving on taking the drug, MC was actually the first synthetic antimalarial successfully used in SLE.10 The first series of patients with lupus treated with MC was published by Page in The Lancet in 1951.11 18 patients with cutaneous lupus, 1 of them with acute systemic disease and 2 with concomitant rheumatoid arthritis (sic.), were treated with MC, 17 of them showing clinical response. Despite its efficacy, the frequent yellowish skin staining, the reports of (rare) cases of aplastic anaemia and, overall, the availability of HCQ after 1955, led to the almost complete disappearance of MC from the therapeutic armamentarium of SLE,12 being relegated to those patients with ocular toxicity due to HCQ (or chloroquine), since MC does not cause retinopathy.1 9 However, this lack of macular toxicity makes MC not only a good substitute for other antimalarials in patients at risk of or with proven ocular toxicity, but also a potential option for combination therapy with HCQ, with synergistic immunomodulatory effects and not increasing the risk of retinal damage.9 In 2000, Toubi et al reported a case series of six patients in whom MC was added to HCQ to treat active SLE with cutaneous and/or articular involvement, five of them with other organ manifestations, including kidney disease in two patients.13 In 2018, our group published a cohort study including 46 patients from two academic Spanish centres (Hospital Universitario Cruces and Hospital Clínico San Cecilio), showing a complete/partial response rate according to the clinicians’ evaluation over 90%.4 Since then, MC has become a central agent for the management of cutaneous, articular and/or serosal SLE in our unit. In the most recent European Alliance of Associations for Rheumatology recommendations, MC is mentioned as an alternative for cases of toxic retinopathy or as an add-on to HCQ, only in patients with skin involvement.3 The authors remark that MC use is limited by frequent intolerance and unavailability in many countries.3 Only guidelines for cutaneous lupus erythematosus recommend the addition of MC to HCQ before introducing IS.14 Therefore, the use of MC in non-cutaneous lupus is not usually considered. Recent studies have evaluated the differential profile of MC compared with HCQ. While both antimalarials exert their effects by inhibiting toll-like receptors 7/9,12 15 MC has exhibited some alternative properties. It is able to reduce B-lymphocyte activating factor (BLyS), as shown by Toubi et al in a study involving 29 patients who received MC as an add-on therapy to treatment with HCQ and AZA.16 The reductions in BLyS, anti-dsDNA and anticardiolipin levels were more marked among patients with higher baseline SLEDAI scores. Alves et al showed a significant effect of MC in suppressing tumour necrosis factor (TNF)-α, in addition to interferon-α (IFN-α), in cutaneous lupus and dermatomyositis, while HCQ was only able to suppress IFN-α.17 This reduction in TNF-α could be explained by the ability of MC to inhibit the nuclear factor-κβ pathway.18 Although TNF-α is not a cytokine mainly involved in the pathogenesis of SLE, several studies have remarked that patients with SLE have higher circulating levels when compared with healthy controls.19 20 A recent study has found that patients with cutaneous lupus and a high population of myeloid dendritic cells with high TNF-α expression are more likely to be refractory to HCQ therapy, and on the contrary, more likely to have a better response to MC.15 Moreover, an observational study by Cortes-Hernandez et al showed the efficacy of etanercept in refractory arthritis in patients with SLE,21 a fact that could support an additional effect of MC also in patients with articular disease, as shown in this study. In this study, we confirm the beneficial effects of MC in mild–moderate active SLE outlined in our previous work.4 Besides expanding the cohort to 106 cases, we also included a validated global outcome criteria such as DORIS remission.8 In a cohort in which almost 80% of patients had not cutaneous disease as the main reason to incorporate MC, 71% patients achieved the main outcome. Such a high response rate is in fact in the range of the 73% excellent or improved response calculated by Wallace in a comprehensive review of 20 studies on MC in lupus (mainly with cutaneous disease and in monotherapy) published in 1989.9 As shown further in this section, the frequent use of MP and the reduced doses of GCs (of note, one of the items of DORIS remission is prednisone dose ≤5 mg/day) may have also had an impact on these results. This remarks on the concept of treating SLE from a multitarget approach, as recommended in the most recent guidelines3 and supported by the high remission rates seen in other studies in our cohort.22 23 Moreover, the rates of response did not differ depending on the clinical manifestation causing MC therapy. On the other hand, we found differences in remission rates between patients in whom MC was started as the first drug after GCs and HCQ, who were more likely to respond, and those already receiving IS ± biologics. In any case, more than 60% patients in the latter group also achieved DORIS remission within 6 months after the addition of MC. This means that, although using MC as the first drug after the failure of first-line therapy with GCs and HCQ is a better option, the addition of MC to a multidrug regimen can also be of help in a substantial number of patients. Likewise, a steroid-sparing effect of MC was observed, allowing for a reduction in the mean doses of prednisone from 6 to 4.3 mg/day. More than 95% patients in the cohort were on prednisone ≤5 mg/day within 6 months. On the other hand, we could not confirm higher beneficial effects of MC in smokers, as shown in our previous work.4 Regarding safety, we reported 18 cases of adverse effects (17%), most of them mild, only requiring drug withdrawal in 7 patients. Only one patient suffered from central anaemia, and toxic effects reverted in all patients after the cessation of MC. These results confirm that haematological toxicity, while serious, is very rare. The incidence of aplastic anaemia among World War II soldiers taking MC was very low (0.003%), and such an effect is usually precluded by a characteristic lichenoid rash.24 If the drug is withdrawn at that time, toxicity is completely reversible. In fact, this rare kind of skin toxicity was the cause of MC withdrawal in one of our patients, without subsequent haematological involvement. Therefore, our experience contradicts the statement that the use of MC is limited by the frequent intolerance.3 We acknowledge a number of limitations of our study. This is an observational cohort, without a control group other than patients themselves. However, the whole group presented with active disease despite therapy including, at least, HCQ and prednisone, and the rate of DORIS remission was high enough as to validate the efficacy of MC. Due to the low prevalence of severe disease within this cohort, the benefit of MC therapy was mainly shown in patients with mild–moderate disease, a fact that should be taken into account when making clinical decisions. The low baseline doses of prednisone (6 mg/day, similar in both subcohorts) could be an object of criticism regarding the definition of refractoriness. However, we believe that this adds further value to our results, remarking that the use of combination therapy can induce rapid remission of active SLE without the need for exposing patients to potentially harmful doses of prednisone.22 23 The concomitant use of MP was common in both groups, but significantly more frequent in the ST group (table 1). Although we could not identify an additional independent effect of MP in achieving DORIS remission within this cohort, this effect cannot be excluded in view of the recent results of the Lupus-Cruces-Bordeaux cohort.23 So, although the whole beneficial effect could not be only attributed to MC (in both groups), the persistence of remission at 6 months makes it likely that MC played an important role. In any case, the concomitant use of MC and MP offers, again, a good alternative to increasing oral prednisone doses. Moreover, MC was mostly used in combination with HCQ 200 mg/day, the usual dose in our unit.25 While a big debate regarding the optimum dose of HCQ is still alive,25 recent data using the Kaiser Permanente Northern California integrated health network support the idea that the risk for HCQ toxicity continues to decrease when lowering the dose below 5 mg/kg/day.26 Thus, we believe, recalling an editorial of McCune and Gonzalez-Rivera in 2015,27 that adding MC is a more rational option than increasing the dose of HCQ and taking a higher potential risk of macular toxicity, given the overwhelming evidence supporting the extensive beneficial effects of HCQ in lupus, beyond controlling disease activity.2 Every effort should be made in order to assure the very long-term therapy of every patient with SLE with this essential agent.3 Additional limitations include that most treated patients were white and had universal and easy access to public and free health facilities. Finally, there is a current shortage of MC in many countries, being difficult to obtain in the USA.28 In Spain, it is available as a foreign medication, with irregular availability throughout the different regions of the country. In our centre, patients can collect the pills free of charge from the hospital pharmacy, which makes it very easy for us to prescribe MC. In conclusion, this study constitutes the largest world cohort of patients with SLE treated with MC + HCQ. Our results support the use of MC as an early add-on therapy to HCQ in patients with active SLE at cutaneous, articular and/or serosal levels, although patients with more complex therapeutic schemes can also benefit from this safe and well-tolerated drug. As an additional advantage, MC can allow the use of lower doses of prednisone and, also, of HCQ, with the implications in preventing ocular toxicity in the long run. Our results could stimulate further prospective trials that would hopefully validate and better define the efficacy of MC in SLE and cutaneous lupus erythematosus and, so, increase its production, availability and affordability worldwide.28 Footnotes Funding: GR-I was supported by the Department of Education of the Basque Government, research grant IT 1512-22. Patient consent for publication: Not applicable. Ethics approval: This study involves human participants and was approved by CEIC Hospital Universitario Cruces. CEIC E08/35. Participants gave informed consent to participate in the study before taking part. Provenance and peer review: Not commissioned; externally peer reviewed. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Data availability statement Data are available upon reasonable request. References Associated Data Data Availability Statement Data are available upon reasonable request.