Atopic dermatitis is a chronic inflammatory disease of childhood that is characterized by recurrent outbreaks of very pruriginous lesions. Some children have severe disease that is refractory to conventional treatment with topical corticosteroids, thus necessitating immunosuppressive therapy to achieve control.1 Azathioprine is a prodrug that is converted into 6-mercaptopurine after absorption in the intestine. 6-Mercaptopurine is subsequently metabolized in the liver and gastrointestinal tract by 3 main enzymes: hypoxanthine-guanine phosphoribosyltransferase, xanthine oxidase, and thiopurine methyltransferase. Azathioprine antagonizes purine metabolism and thus inhibits the synthesis of DNA, RNA, and proteins, as well as cell mitosis.2 It is used in dermatology as a corticosteroid-sparing agent in a wide variety of inflammatory and autoimmune diseases, including (off-label) atopic dermatitis and nummular eczema.3–5 Its most common adverse effects are myelosuppression, liver toxicity, gastrointestinal symptoms, hypersensitivity reactions, and pancreatitis. Measurement of the activity of the enzyme thiopurine methyltransferase in serum, which varies depending on specific gene polymorphisms,1 helps to prevent some of these adverse effects, mainly myelosuppression. Given that the agent is not authorized for infantile atopic dermatitis, it is reserved for specific cases where topical treatments or other types of systemic treatment are not effective.

To evaluate the efficacy and tolerance of azathioprine in children with severe atopic dermatitis that is refractory to conventional approaches.

We performed a single-center retrospective observational cohort study of 11 patients with severe atopic dermatitis treated with azathioprine at Hospital Infantil Universitario Niño Jesús, Madrid, Spain between January 2007 and May 2017. The local ethics committee approved the study. Patients were identified using computerized and imaging databases. Atopic dermatitis was defined using the Investigator Global Assessment scale, which defined the disease as the presence of intense, dark erythema with induration, excoriated papules, exudation, and crusting.6,7 Before receiving azathioprine, all patients had had a poor response to standard treatment with moderate-potency topical corticosteroids (class 3) applied twice daily or once daily with wet-wrap dressing techniques and/or topical immunomodulators. No patients received oral azathioprine as their first line of immunosuppressive therapy: 10 had previously received ciclosporin A, and 1 had received methotrexate. Patients were allowed to apply moderate-potency topical corticosteroids during treatment with azathioprine. In all cases, levels of thiopurine methyltransferase were assessed before starting treatment, and the dose of azathioprine was adjusted according to outcome (< 5.1U/mL red cells, not administered; 5.1-13.7U/mL, 0.5mg/kg/d; 13.8-18.0U/mL, 1.5mg/kg/d; 18.1-26.0U/mL, 2.5mg/kg/d; and 26.1-40U/mL, 3.0mg/kg/d).

The demographic data recorded included sex, age at onset of atopic dermatitis, age at initiation of treatment, history of asthma and/or sensitization to food (defined as specific immunoglobulin [Ig] E levels for food greater than 95% of the positive predictive value), total IgE levels, and eosinophilia in blood (defined as peripheral eosinophilia> 500/μL). Drug-related variables included the initial dose of azathioprine, dose increases (yes/no), and duration of treatment. Outcome was assessed based on the clinical response after 4 weeks of treatment, after 3-4 months of treatment, and at>6 months of treatment and on long-term efficacy (time to relapse and disease-free period after withdrawal of azathioprine). Given that the Investigator Global Assessment scale is difficult to evaluate in a retrospective study, evaluation was based on the clinical notes of the attending dermatologist, with outcome classified as complete or near-complete clearance (improvement≥90%), marked improvement (50%-89%), slight improvement (<50%), and failure (no improvement). In the case of good to excellent outcome (complete clearance, near-complete clearance, or marked clearance), the disease was considered to be in remission when the patient's condition had resolved and could be managed with topical therapy (moderate-potency corticosteroids/topical immunomodulators) and to be recurrent when flares made it advisable to administer another systemic immunosuppressive drug.

Laboratory monitoring consisted of a complete blood count and biochemistry before initiation of treatment, 15 days after initiation, 4 weeks after initiation, and every 3 months thereafter. Clinical follow-up was at least at the same intervals. The assessment of adverse effects included symptoms (eg, asthenia), gastrointestinal adverse effects (eg, nausea, vomiting, and abdominal pain), physical examination, complete blood count, serum urea, creatinine, and serum liver enzymes. Any complaint that could be considered a clinical adverse effect of azathioprine was also reported.

We reviewed 11 patients (6 boys, 5 girls) with a mean age of 13 years (range, 8-18 years). Mean age at initiation of treatment was 10.9 (2.2) years (95% CI, 8.6-13.1 years). The mean initial dose was 1.8 (0.2) mg/kg/d. We evaluated the response to treatment at 4 weeks, between weeks 12 and 16, and from 6 months onward. One patient remains in follow-up.

Table 1 shows demographic data and drug-related variables. Two patients had to suspend therapy with azathioprine because of adverse effects, one during the first month and the other at the fourth month after starting treatment. In the group of patients who completed treatment with azathioprine, the mean duration of therapy was 10.8 (5.7) months (95% CI, 5.0-16.6). After 1 month of treatment, no patients had a complete or near-complete response. Between 12 and 16 weeks, clearance was near-complete in 4 of 9 patients, a marked improvement was observed in 3 patients, and a slight improvement in 2. Two of the 3 patients who had improved at the fourth month continued to improve until near-complete clearance was observed at the sixth month of treatment. Therefore, of the 9 patients whose treatment was extended for a minimum of 6 months, clearance was almost complete in 6 at this point.

We found no statistically significant association between response to treatment and a history of allergic sensitization, history of asthma, eosinophilia, and IgE levels.

The mean dose of azathioprine used in the present study was 1.8 (0.2) mg/kg/d (95% CI, 1.5-2.1). The mean follow-up time was 33.1 (12.0) months. One patient was still taking azathioprine at the end of the study. After suspension of treatment, 2 patients with complete clearance required new systemic therapy at 8 and 20 months, respectively. The remaining cases were managed with topical corticosteroids, immunomodulators, or both.

Adverse effects were recorded in 3 patients (27%). One patient experienced nausea and epigastric pain, one experienced a mild increase in transaminases, and a third patient experienced a mild transitory increase in transaminases that resolved spontaneously. In the case of the 2 patients with epigastric pain, treatment was suspended owing to the intensity of the pain. No patients had a severe intercurrent infection or complained of effects other than those already commented upon.

Although azathioprine is not authorized by the United Stated Food and Drug Administration or the European Medicines Agency for the treatment of atopic dermatitis in children, it has proven beneficial in 2 randomized placebo-controlled trials in adults with atopic dermatitis treated with azathioprine for 12 weeks and in several retrospective pediatric case series.4,8–10 Both our results and those of previous studies1,10 that evaluated the efficacy and tolerability of azathioprine in children with severe atopic dermatitis indicate that most patients have a good or very good response at 12-16 weeks after starting treatment. Thus, in our study, clearance was complete or near-complete in 7 of 9 patients; this figure is slightly higher than that reported elsewhere.9Table 2 shows studies that analyzed treatment with azathioprine in pediatric patients with atopic dermatitis. The main disadvantage of azathioprine is not its efficacy, which is similar to that of other immunosuppressants (eg, ciclosporin and methotrexate),11,12 but the time to response, since it does not take effect for at least 3 months.10 In fact, consistent with the view of other European specialists,13,14 this interval and the adverse effects profile meant that azathioprine was not our first choice for any of the patients. We also agree with other authors that the adverse effects we observed were mainly mild gastrointestinal adverse effects, which led to suspension of treatment in only 2 cases.5,10 It is noteworthy that in both cases, these effects appeared during the first month of treatment and resolved once medication was suspended.

Our study is subject to a series of limitations, mainly its small sample size. In addition, it was a retrospective study, and adverse events may have been underestimated if patients were not actively asked about them. For the same reason, it is difficult to evaluate the Investigator Global Assessment scale. Finally, the limited follow-up period prevents us from drawing definitive conclusions about the duration of remission and long-term safety of treatment with azathioprine.

In conclusion, our data confirm the efficacy and tolerability of azathioprine in the treatment of moderate to severe atopic dermatitis in children; therefore, we should include it as part of our systemic treatment options in these patients. While azathioprine does not seem to have serious adverse effects on children in the short term, strict monitoring is mandatory in this age group.

The authors declare that they have no conflicts of interest.