In total, 12 patients (median age 50.5 years, range 23–60) had started treatment with dupilumab and eight patients (median age 56 years, range 43–81) had started treatment with methotrexate. Patients treated with dupilumab had been given at least one other systemic immunosuppressive treatment before dupilumab was considered: methotrexate (n=12), ciclosporin (n=6), omalizumab (n=2), apremilast (n=3), acitretin (n=1), azathioprine (n=1), and/or alitretinoin (n=1). Half of the patients (n=4) treated with methotrexate had not used any systemic immunosuppressive treatment at baseline. The other half had been treated with omalizumab (n=1), apremilast (n=1), acitretin (n=1), and/or alitretinoin (n=2). During the study period, none of the study participants used systemic glucocorticoids.
All patients were classified as having moderate or severe AD refractory to conventional treatment. Outcome measures for AD at baseline are shown in Table 1. Age, sex, EASI, POEM, VAS for pruritus, DLQI, and weight were comparable between patients with dupilumab and patients with methotrexate at baseline (start of treatment). DLQI was significantly higher among patients with methotrexate (p=0.025), and EASI was significantly higher among patients with dupilumab (p=0.045) (Table 1).
At follow-up, patients treated with dupilumab had gained weight significantly (mean weight change: 6.1kg, range [0.1–18.0], p=0.002), while the weight of patients with methotrexate had not changed (mean weight change: -3.1kg, range [-17.0–2.0], p=0.161), Fig. 1. The majority of the patients showed a positive response to treatment (Fig. 2). Most patients treated with dupilumab had improved all outcome measures (EASI, POEM, VAS, DLQI, and MADRS) significantly. Moreover, at follow-up at 6, 9, or 12 months, they reached EASI-90 (n=6), EASI-75 (n=4), or EASI-50 (n=1), and had a minimal clinically important difference in POEM (n=10, 1 missing value). One patient did not respond to treatment with dupilumab. Patients treated with methotrexate reached EASI-90 (n=3), EASI-75 (n=2), or EASI-50 (n=1) and had a minimal clinically important difference in POEM (n=6). Two patients did not respond to treatment with methotrexate. Treatment was stopped and they switched to dupilumab instead.
Weight change in relation to treatment response
We performed a sensitivity analysis among patients with AD successfully treated with dupilumab (n=11) and methotrexate (n=6), Table 2. Patients successfully treated with dupilumab were comparable with all patients treated with dupilumab (mean weight change: 6.6kg, range [0.1–18.0]). Patients treated with methotrexate had a significant response to treatment (EASI, VAS, DLQI, and MADRS were significantly reduced) and tended to lose weight from baseline to follow-up (mean weight change: -4.3kg, range [-17.0–2.0]), but not significantly.
Appetite and sleep disturbance due to itching in relation to weight change
At baseline, the majority of patients treated successfully with dupilumab and methotrexate reported normal or increased appetite (7 of 11 and 3 of 5 [1 missing data], respectively). The other patients reported slightly reduced appetite (1 or 2 on scale 0–6). No patient reported no appetite or need persuasion to eat (≥3 on scale 0–6) (Additional files 1 and 2). In the dupilumab group, 3 of 4 patients still reported slightly reduced appetite at follow-up, and one patient reported normal or increased appetite. Among patients treated with methotrexate, 2 of 2 patients were improved and reported normal or increased appetite at follow-up. In the visual assessment, there was no correlation between reported appetite and weight change among patients treated with dupilumab or methotrexate.
Disturbed night sleep Every day last week (4, scale 0–4) was reported at baseline among 6 of 10 (1 missing data) patients treated successfully with dupilumab. At follow-up, one patient still reported disturbed night sleep Every day last week, and all others (n=9) reported disturbed night sleep No days (0, scale 0-4). In all, 5 of 10 clearly improved their night sleep from baseline to follow-up. For five patients, night sleep was unchanged (one patient reported disturbed night sleep Every day, and the other four reported No days). Both patients with improved night sleep (n=5) and those with unchanged night sleep (n=5) gained weight (mean weight change: 7.2kg and 3.8kg, respectively) with no significant difference between the groups (p=0.076). Among patients treated successfully with methotrexate, all reported disturbed night sleep in the last week at baseline (1–2 days [n=1], 3–4 days [n=1], 5–6 days [n=2), Every day [n=2]). All improved their night sleep from baseline to follow-up; two patients who reported Every day at baseline reported 1–2 days and 3–4 days, respectively, at follow-up. All others reported No days at follow-up. Patients treated successfully with methotrexate and who improved their night sleep during treatment did not gain weight (mean weight change: -4.3kg) (Additional files 1 and 2).