Patients with atopic dermatitis with filaggrin loss-of-function mutations show good but lower responses to immunosuppressive treatment

Dear Editor, Filament‐aggregating protein (filaggrin) mutations are a strong risk factor for developing atopic dermatitis but the relationship has not been thoroughly studied. To investigate whether filaggrin mutations influence the outcome of immunosuppressive treatment we studied patients with severe atopic dermatitis in a single‐blinded randomized controlled trial with methotrexate or azathioprine during a 24‐week regimen.1 Buccal mucosa swabs were collected 2 years after randomization from 36 of the 42 participants in the randomized controlled trial (86%) to determine their filaggrin genotype status (R501X, 2282del4, R2447X, S3247X and 3321delA mutations). In the original trial both methotrexate and azathioprine induced significant reductions in the Scoring of Atopic Dermatitis index (SCORAD) at 12 and 24 weeks, with no significant difference found between the treatments.1 We therefore pooled all patients together for this substudy and divided them into a filaggrin mutation group (FMG) and a non‐FMG. The primary outcome parameter was the SCORAD, which was obtained at baseline and weeks 2,4,8,12 and 24, and analysed as repeated measurements (lme4 package in R, version 3.2.5, https://www.r-project.org). All SCORAD values were obtained from patients who had been receiving systemic immunosuppressive treatment with or without topical immunosuppressive treatment, except for three values in three patients. Two patients stopped systemic treatment after experiencing adverse events (FMG, week 2; non‐FMG, week 4) and their last scores, obtained within 2 weeks of stopping systemic therapy, were for topical immunosuppressive treatment alone. One (FMG) patient stopped systemic treatment at week 12, and the values at week 24 were for patients undergoing only topical immunosuppressive treatment. As the patients’ mean age was significantly higher in the FMG than in the non‐FMG (P < 0·01, Table S1; see Supporting Information), all analyses were corrected for age as a possible confounder, and age was included in the models as a potential confounder variable. As the treatment regimen (azathioprine or methotrexate) did not influence treatment success in the FMG compared with the non‐FMG, we included in the mixed models as covariates visit, mutation, age and an interaction term for visit * mutation. The main limitation of this substudy was the small sample size and our subsequent inability to randomize on filaggrin mutation status.1 Filaggrin mutations were found in 13 of the 36 patients (36%, Table S2; see Supporting Information). Every patient in the FMG and the non‐FMG showed an improvement of SCORAD at week 24 compared with baseline (Fig. S1; see Supporting Information). The mean SCORAD in the FMG decreased by 20·7 points (95% confidence interval 14·7–26·6, P < 0·01) and in the non‐FMG it decreased by 31·3 points (95% confidence interval 26·3–36·2, P < 0·01). However, over the course of 24 weeks the FMG showed less improvement in SCORAD than the non‐FMG (P = 0·02) and the improvement in the SCORAD of patients in the FMG stagnated during the course of treatment (Fig. 1). When the three SCORAD values obtained under topical treatment alone were excluded from data analysis, the FMG showed a trend towards less improvement in the course of 24 weeks than the non‐FMG, as a significant difference was not reached (P = 0·08). Median dosages of topical immunosuppressive treatment (FMG 132 g, non‐FMG 120 g, P = 0·51) methotrexate (FMG 83·8 g, non‐FMG 90 g, P = 0·26) and azathioprine (FMG 837·5 g, non‐FMG 825 g, P = 0·75) were comparable between groups over the course of 24 weeks (Mann–Whitney U‐test). Mean (SD) reduction in Scoring of Atopic Dermatitis index (SCORAD) from baseline to 24 weeks, adjusted for age. Differences in outcomes between the filaggrin mutation group (FMG, n = 13) and non‐FMG (n = 23) when on combined systemic and topical immunosuppressive treatment or topical immunosuppressive treatment alone. All data points are included. Ggplot package in R was used to draw the figures. The grey area represents the 95% confidence interval for the mean reduction in SCORAD in the FMG and non‐FMG. Dark grey represents the overlap between the scores of the groups. These results show that both groups showed an improvement in SCORAD compared with baseline. However, patients in the FMG showed less improvement in SCORAD in the course of 24 weeks than those in the non‐FMG (P = 0·02), and their improvement stagnated during the course of the treatment. It is possible that T‐helper‐cell‐2 cytokines can downregulate filaggrin protein expression2 and, in patients with atopic dermatitis and severe skin inflammation, skin barrier impairment seems similar in patients with and without filaggrin mutations and is correlated with the SCORAD.3 At the initiation of the study, levels of skin barrier impairment and inflammation were probably comparable in the two groups, which is supported by their similar mean SCORAD. When decreasing inflammation by immunosuppressive treatment, downregulation of filaggrin expression by the inflammatory infiltrate will be abrogated, allowing the skin barrier to be restored and, probably, the establishment of normal filaggrin protein levels in the non‐FMG. In contrast, patients with filaggrin mutations will also experience an improvement of their atopic dermatitis due to the immunosuppression but the impairment of the barrier function and reduced filaggrin protein levels will still be present in the end because of their genetically caused filaggrin protein deficiency. This may explain both the lower therapeutic responses in the FMG over the course of 24 weeks and why the initial improvement stagnates during the treatment period. Although both groups showed a significant improvement in SCORAD compared with baseline, filaggrin mutations may have a negative impact on the success of immunosuppressive treatment in severe atopic dermatitis. Additional research investigating the relationship between filaggrin mutation status and treatment success is necessary to confirm these findings. Funding sources: none. Conflicts of interest: W.H.I.M has filed patents on genetic testing and treatment development related to the filaggrin gene. The remaining authors have no conflicts of interest. Table S1. Demographics and baseline characteristics of severe atopic dermatitis patients with and without filaggrin gene mutations. Table S2. Type of filaggrin gene mutations among the 13 filaggrin gene mutation carriers. Figure S1. Overview of SCORAD values measured at baseline (week 0), week 2, 4, 8, 12 and 24 in severe atopic dermatitis patients with (n = 13) and without filaggrin (n = 23) mutations.