Background: Dupilumab is the first biologic available to treat atopic dermatitis (AD). Its effectiveness and safety were demonstrated in clinical trials. Objective:We sought to assess the effectiveness and safety of dupilumab in adults with AD in a real-life French multicenter retrospective cohort.Methods: Patients were included between March 2017 and April 2018. Efficacy outcomes were collected both at baseline and three months (M3), when available, including SCORAD (Scoring Atopic Dermatitis) and EASI (Eczema Area and Severity Index) scores. Adverse events (AE) were recorded at the follow-up. Results:We included 241 patients. The median follow-up time was 3.8±3.7 months. SCORAD75 and EASI75 were achieved in 27/163 (16.6%) and 40/82 (48.8%) patients, respectively. The median SCORAD and EASI at M3 were significantly lower compared with baseline (25±21 vs 56±27.4, p<10 -9 and 4.1±6.8 vs 17.9±15.4, p<10 -9 , respectively).Conjunctivitis was reported in 84/241 (38.2%) patients. The proportion of eosinophilia (>500/mm 3 ) during follow-up (57%) was higher than at baseline (33.7%) (n=172, p<10 -6 ).Dupilumab was stopped in 42 cases, 27 of which were due to an AE. Limitations:No control group, missing data. Conclusion:This real-life study demonstrated results similar to clinical trials, with regard to dupilumab effectiveness, but revealed a higher frequency of conjunctivitis and eosinophilia.

Dipeptidyl peptidase-4 inhibitors have been suspected to induce bullous pemphigoid (BP). The objective of this study was to compare the observed frequency of gliptin intake in a large sample of 1,787 BP patients diagnosed between 2012 and 2015 in France, with the expected frequency after indirect age standardization on 225,412 individuals extracted from the database of the National Healthcare Insurance Agency. The secondary objective was to assess the clinical characteristics and the course of gliptin-associated BP, depending on whether gliptin was continued or stopped. The observed frequencies of intake of the whole gliptin class and that of vildagliptin in the BP population were higher than those in the general population after age standardization (whole gliptin class: 6.0%; 95% confidence interval ¼ 4.9e7.1% vs. 3.6%, observed-to-expected drug intake ratio ¼ 1.7; 95% confidence interval ¼ 1.4e2.0; P < 0.0001; vildagliptin ¼ 3.3%; 95% confidence interval ¼ 2.5e4.1% vs. 0.7%, ratio ¼ 4.4; 95% confidence interval ¼ 3.5e5.7; P < 0.0001). The association of any gliptinþmetformin was also higher than in the general population, ratio ¼ 1.8 (95% confidence interval ¼ 1.3e2.4; P < 0.0001). Gliptin-associated BP had no specific clinical characteristics. Gliptin was stopped in 48 (45.3%) cases. Median duration to achieve disease control, rate, and delay of relapse were not different whether gliptin was stopped or continued. This study strongly supports the association between gliptin intake, particularly vildagliptin, and the onset of BP.

AIMSLinear immunoglobin A (IgA) bullous dermatosis is a rare autoimmune dermatosis considered spontaneous or drug-induced (DILAD). We assessed all DILAD cases, determined the imputability score of drugs and highlighted suspected drugs. METHODSData for patients with DILAD were collected retrospectively from the French Pharmacovigilance network (from 1985 to 2017) and from physicians involved in the Bullous Diseases French Study Group and the French Investigators for Skin Adverse Reactions to Drugs. Drug causality was systematically determined by the French imputability method. RESULTSOf the 69 patients, 42% had mucous membrane involvement, 20% lesions mimicking toxic epidermal necrolysis (TEN), 21% eosinophil infiltrates and 10% keratinocytes necrosis. Direct immunofluorescence, in 80%, showed isolated linear IgA deposits. Vancomycin (VCM) was suspected in 39 cases (57%), 11 had TEN-like lesions, as compared with three without VCM suspected. Among the 33 patients with a single suspected drug, 85% had an intrinsic imputability score of I4. Among them, enoxaparin, minocycline and vibramycin were previously unpublished. For all patients, the suspect drug was withdrawn; 15 did not receive any treatment. First-line therapy for 31 patients was topical steroids. Among the 60 patients with British Journal of Clinical Pharmacology Br J Clin Pharmacol (2019) 85 570-579 570 available follow-up, 52 achieved remission, 10 without treatment. Four patients experienced relapse, four died and five had positive accidental rechallenges. CONCLUSIONSThere is no major clinical difference between DILAD and idiopathic linear IgA bullous dermatosis, but the former features a higher prevalence of patients mimicking TEN. VCM, suspected in more than half of the cases, might be responsible for more severe clinical presentations. We report three new putative drugs. WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT• Linear IgA bullous dermatosis (LABD) is a rare autoimmune dermatosis that may be spontaneous or drug induced.• Vancomycin (VCM) is the most frequent putative drug.• Drug-induced LABD (DILAD) may be more severe than idiopathic LABD, mimicking toxic epidermal necrolysis with Nikolsky sign and large erosions WHAT THIS STUDY ADDS• In this large series of DILAD, we systematically assessed the imputability of suspect drugs. Most of them are antibiotics with VCM as the most frequent. Eighty percent of suspect drugs had a high causality score. Three drugs had never been reported to induce DILAD before (enoxaparin, minocyclin, vibramycin). • Twenty percent of cases (especially those induced by VCM) mimicked toxic epidermal necrolysis. Data collectionFor each case, data collected from pharmacovigilance or medical files included epidemiological and clinical characteristics Drug-induced linear immunoglobulin A bullous dermatosis Br J Clin Pharmacol (2019) 85 570-579 571 Drug-induced linear immunoglobulin A bullous dermatosis Br J Clin Pharmacol (2019) 85 570-579 579