A comprehensive database search was performed individually by the members of the working group. They performed computerized database searches of Medline (accessed by PubMed) and Embase for articles published between January 1, 2005, and December 31, 2014, using combinations of "atopic eczema", "atopic dermatitis", "antihistamine", "antimicrobial, "antifungal", "antiviral", "corticosteroids", "cyclosporine", "azathioprine", "methotrexate", "mycophenolate mofetil", "biologics", "interferon-γ", "alitretinoin", "immunoglobulin", "thymopentin", "allergen-specific immunotherapy", "phototherapy", "complementary and alternative medicines", "probiotics", "prebiotics", "vitamin D", "essential fatty acid", "herb medicine", and "acupuncture". The searches were supplemented by manual searches of references from relevant systematic reviews and guidelines of other groups. The members collected all relevant statements relating to AD management.

The members of the working group graded the evidence and then classified the strength of recommendation for each statement. The evidence for each statement was graded as follows: level 1, systematic review of randomized controlled trials (RCTs) or individual RCT; level 2, systematic review of cohort studies and individual cohort study (including low-quality RCT); level 3, systematic review of case-control studies and individual case-control study; level 4, case series (and poor-quality cohort and case-control studies); and level 5, expert opinion. The strength of recommendation was classified as A (level 1), B (level 2 and 3), C (level 4),or D (level 5) (Table 1)2.

Fifty-four council members of the KADA were asked to provide their level of agreement with each draft statement, using a voting scale of 1~9 (where 1 denotes strong disagreement and 9 denotes strong agreement). Thirty-nine Korean experts participated in the vote. Each voting score was allocated to one of three groups: 1~3 (disagreement), 4~6 (neutrality), and 7~9 (agreement). Consensus was defined as ≥75% of participants providing a score within the 7~9 range (agreement). Consensus recommendations were derived after three rounds of voting.

AD is characterized by itching as a subjective symptom. Sedating and non-sedating antihistamines have been used for decades to treat AD. There is insufficient evidence to recommend the general use of either type of antihistamine in AD treatment, as other pruritogenic substances and histamines contribute to pruritus-related AD3. However, Korean experts recommend the use of antihistamines in attempts to relieve pruritus and prevent exacerbation due to scratching in patients with mild to severe AD (Table 2)456789.

Short-term, intermittent use of sedating antihistamines, such as hydroxyzine and chlorpheniramine, may be beneficial when there is sleep loss due to itching3.

An RCT confirmed that the addition of fexofenadine to a topical corticosteroid (TCS) reduces pruritus associated with AD4. The long-term use of cetirizine in infants with severe AD had TCS-sparing effects, which were used as an indirect measure of the efficacy of cetirizine in treating pruritus5. Non-sedating antihistamines may be helpful, particularly when the patient has comorbidities such as bronchial asthma, rhinoconjunctivitis, or urticaria56.

General recommendations for antihistamine selection and dosing regimens (dosage, and continuous vs. intermittent administration) have not been established and treatment should consider individual factors.

Common side effects of antihistamines include undesired sedation, even with non-sedating formulations, and anticholinergic symptoms, such as dry mouth, blurred vision, and tachycardia3. In general, the long-term use of antihistamines is safe. No laboratory monitoring is required. If cardiac toxicity is suspected, an electrocardiogram should be obtained to assess dysrhythmia3.

Having an impaired skin barrier, patients with AD are likely to develop various secondary infections, including Staphylococcus, herpes simplex, molluscum contagiosum, and Malassezia furfur fungal infection.

Although S. aureus can be cultured from the skin of an estimated 5% of the population without AD, this microbe has been isolated from more than 90% of adult AD patients310. The clinical relevance of bacterial overgrowth is patient-dependent, so the use of systemic or topical antibiotics to treat non-infected AD is not recommended. Short-term treatment with topical or systemic antibiotics may be beneficial in addition to standard, appropriate treatment if the skin is obviously superinfected with bacteria (1a, A)31011. In particular, the continuous use of antibiotics, regardless of whether they are topical or systemic, should be avoided to reduce the risk of bacterial resistance (Table 3)311. Bacterial culture with antibiotic susceptibility profiling may be appropriate for recurrent or non-responsive skin infections3.

In patients with moderate to severe AD and clinical signs of secondary bacterial infection, bleach baths and intranasal mupirocin may be recommended to reduce disease severity12.

Eczema herpeticum occurs more frequently in AD patients than in normal individuals and has a tendency to disseminate. Patients with severe AD, untreated skin lesions, early disease onset, and high total serum immunoglobulin E (IgE) levels are at high risk of developing eczema herpeticum. Pretreatment with TCSs does not imply higher risk13. Eczema herpeticum should be treated without delay using systemic antiviral therapy.

Secondary infections with yeasts have also been implicated as trigger factors in AD14.

Therefore, systemic and topical antifungal agents have been proposed for treatment of "head and neck" dermatitis in AD (Table 3)1516.

Systemic immunomodulators can be used in AD patients who show an inadequate response to conventional topical agents or phototherapy-which may have a negative impact on sleep, emotional stress, or social activities-or in patients with a SCORing AD (SCORAD) index of 40 or more17. Systemic immunomodulators, such as corticosteroids, cyclosporine, methotrexate (MTX), mycophenolate mofetil (MMF), and azathioprine (AZP), are used in clinical settings. In contrast, there are only limited reports on leukotriene inhibitors, oral calcineurin inhibitors, and interferonγ (IFN-γ). Further investigations are necessary before these options can be used in routine treatment. Aside from corticosteroids, all systemic immunomodulators may increase the risk of photocarcinogenesis when used together with phototherapy; this combination is therefore not recommended. Live attenuated vaccination is usually contraindicated when systemic immunomodulators are used. The dosage and material in the vaccine should be carefully examined when vaccine use is considered. Recommendations for systemic immunomodulators are summarized in Table 435181920.

Corticosteroids are natural products of the adrenal gland that are used to regulate the immune system and stress response in humans. Although systemic corticosteroids dramatically improve the clinical symptoms of AD, their administration should generally be avoided because of adverse effects and the rebound phenomenon. Rebound flare is frequently observed after the abrupt cessation of systemic corticosteroids. Increased production of IgE by B cells in AD patients has been reported after treatment with oral prednisolone2122. Once clinical improvement has been achieved, it is very important to taper the dosage gradually over time to minimize the likelihood of a rebound effect.

Clinical trials have shown that corticosteroid concentrations in the skin following the administration of a potent TCS (clobetasol propionate 0.05%, hydrocortisone 2.3%, or triamcinolone 0.1%) are similar to those achieved with medium doses of oral prednisone21. If the skin is severely damaged, however, the distribution of topical treatments is extremely irregular and oral administration is safer and more controllable. In all other situations, TCSs are the preferred option.

Continuous or chronic intermittent use of systemic corticosteroids in AD is discouraged. However, acute usage may be considered as a transitional therapy in severe, rapidly progressive, or debilitating cases during the initiation of treatment with nonsteroidal systemic immunomodulatory agents that have more favorable side-effect profiles, or phototherapy3. Some clinicians argue that systemic corticosteroids can be used safely for up to six weeks in combination with TCSs or topical calcineurin inhibitors23.

Dosage is based on body weight, ranging from 0.5 to 1.0 mg/kg per day during acute flares24. Significant adverse effects of the chronic use of systemic corticosteroids include hypertension, diabetes, glucose intolerance, gastritis, weight gain, osteoporosis, skin atrophy, glaucoma, Cushing's syndrome, and emotional lability3. Children and adolescents receiving systemic steroids continuously may exhibit decreased linear growth while taking the medication25.

Cyclosporine is the primary choice for systemic immunomodulators in moderate to severe AD patients who are unresponsive to topical therapy and oral antihistamines. The effects of treatment appear two weeks after initiation, with 50%~60% improvement expected in 6~8 weeks26. However, symptoms may manifest themselves again within 8~12 weeks of termination of medication2728.

Cyclosporine can be used in children older than two years of age. Long-term safety in children has not yet been established and caution should be exercised, though many studies have shown that it can be relatively safe in young children2729.

The dosage is commonly started with 2.5 mg/kg/day and increased by 0.5~1.0 mg/kg/day at 2- to 4-week intervals, up to 5 mg/kg/day. Compared to this low dose, faster induction can be achieved by starting treatment with a high dose relative to body weight (5 mg/kg/day) and reducing the dose by 0.5~1.0 mg/kg/day every two weeks based on the clinical response18. There is a report that microemulsion formulations have faster effects than usual formulations30. The effects of treatment appear two weeks after initiation, a relatively fast induction rate compared to other systemic immunomodulators.

The maximum duration for medication has not yet been established, but cyclosporine can be used safely for about 1~2 years31. Common and important side effects include nephrotoxicity, hypertension, tremors, headaches, paresthesia, hypertrichosis, gingival hyperplasia, gastrointestinal (GI) symptoms (nausea, vomiting, diarrhea), flu-like symptoms (myalgia, fatigue), hypertriglyceridemia, electrolyte imbalance (hypomagnesemia, hyperkalemia), jaundice, and susceptibility to infection19. Routine follow-up examinations are required before and after administration. Blood pressure should be measured at every visit. Laboratory testing should be performed upon initiation of medication and every 2~4 weeks for several months as the drug dosage is being increased. During the long-term maintenance period, laboratory testing should be performed at least once every three months (Table 5)3. The drug dosage should be lowered when blood creatinine rises by more than 25% and the patient should be closely monitored to decide whether to stop or continue the medication3. Long-term use of cyclosporine raises the possibility of skin cancer and lymphoma19.

Considering the risk and benefits of AZP, it can be used in patients with moderate to severe AD who do not respond to primary treatment modalities. However, as AZP may cause more severe side effects than cyclosporine and is not as effective, it should be considered as a secondary choice for systemic immunomodulator therapy in adult patients who are unresponsive to or experience side effects with cyclosporine520.

The dosage range of AZP is 1~3 mg/kg/day, usually starting with 1.5 mg/kg/day and increasing by 0.5 mg/kg/day at every visit if the disease does not improve by 25%, up to 2.5 mg/kg/day (Table 6)332.

Myelosuppression is the common and important side effect. Recently, side effects such as skin cancer, T-cell lymphoma of the liver and spleen, and progressive multifocal leukoencephalopathy have raised issues regarding AZP use. Caution should be exercised in long-term use1933.

Because of hematological side effects, the dosage should be established after determining the level of thiopurine methyltransferase (TPMT) in the blood34. However, since the test is only available in a few facilities in Korea and is relatively expensive, it is difficult to apply this test to all patients. To reduce the risk of myelosuppression, the drug dosage should be increased when TPMT levels are high and reduced when TPMT levels are low.

One study reported the use of AZP in children aged two years and older, but the results of one study are not sufficient to recommend AZP use in children35.

MTX is considered as a second-line systemic immunomodulator therapy in adult patients who are unresponsive to or experiencing side effects with cyclosporine. MTX has a favorable risk-benefit ratio, considering the clinical experience in psoriasis patients, and could be used in the long-term maintenance treatment of AD. Patient compliance is relatively satisfactory, thanks to the once-weekly administration regimen. However, further investigations are needed to determine the dose and effects of MTX.

Dosage usually does not depend on body weight. MTX is usually administered in a dosage of 7.5~25 mg per week, starting with 10 mg per week via oral administration and increasing by 2.5~5 mg after every visit, if improvement is less than 25%, up to 25 mg/week (Table 7)332. The same amount can be divided into three doses at 12-hour intervals. In addition to oral administration, MTX can also be given by intramuscular or subcutaneous injection. Injection can lead to high bioavailability, but compliance is low due to the invasiveness of the technique.

Patients can expect a 40%~50% improvement after 12 weeks of MTX use, with one report showing an effect similar to AZP and cyclosporine2432. The maximum effect is usually achieved after about 10 weeks.

Common side effects of MTX include GI symptoms, which can be reduced by switching from oral medication to injection19. The most significant side effects include liver cirrhosis (liver toxicity), myelosuppression, and pulmonary fibrosis. The need for routine liver biopsies is controversial, as the incidence of liver cirrhosis is very low and it is not easy to perform liver biopsies for the cumulative doses of MTX used in the treatment of dermatologic disorders. A test for procollagen type III amino-terminal peptide can be used as a substitute for liver biopsy, but is not applied in Korea. Myelosuppression is reversible once drug administration has been stopped or reduced. Studies have reported pulmonary fibrosis in patients on low-dose regimens and MTX is not recommended in patients with asthma or chronic coughing. Folic acid (1 mg daily) can be added to reduce the incidence of myelosuppression and GI symptoms.

Guidelines for use in children with AD have not yet been established, because of the lack of studies. However, studies and clinical experience in children with psoriasis suggest that MTX can also be expected to be safe for use in children with AD17.

MMF can be used as a second-line systemic immunomodulator therapy in adult patients with severe AD who are unresponsive to or experiencing side effects with cyclosporine, especially in long-term maintenance therapy18.

Clinical improvement using MMF begins after 4~8 weeks, which is slower than cyclosporine, and some patients may experience symptom aggravation in the early phases of treatment20. MMF is known to have a therapeutic effect similar to cyclosporine. However, the effects of MMF are more durable than those of cyclosporine and one study showed effective maintenance four months after discontinuation of the drug36.

MMF therapy can be started at 0.5 g/day and increased up to 3 g/day, depending on the clinical response. The recommended dosage is 1~2 g/day (Table 8)336. MMF is generally tolerated, although common side effects include GI symptoms, headache, flu-like symptoms, and fatigue. Serious side effects, such as leukocytopenia, anemia, thrombocytopenia, or alteration of liver function, are rare compared with other immunomodulators. When administered in doses of 1.5 g/day or less, MMF can be used safely for a long period18. It is known to be relatively safe in children37.

Candidates for ASIT are AD patients whose symptoms are not manageable with proper medication and avoidance measures, those experiencing unacceptable side effects with medication, or those wishing to avoid long-term medication use. A recent meta-analysis provides a moderate level of evidence for the efficacy of ASIT in AD management, although these results are based on a small number of RCTs38. ASIT can be recommended for AD patients with hypersensitivity to house dust mites, pollen, animal allergens, mold or fungi, and hymenoptera39. Appropriate examination of medical history, immediate hypersensitivity skin tests, or tests for serum-specific IgE should be performed before applying ASIT. Currently, house dust mite allergen shows the best therapeutic response to AD treatment using ASIT5. ASIT can be administered by subcutaneous injections (subcutaneous immunotherapy, SCIT) or sublingual drops or tablets (sublingual immunotherapy, SLIT)40. SCIT is effective in the treatment of AD with aeroallergen sensitivity414243. Recently, Novak et al.44 reported the efficacy and safety of SCIT using depigmented polymerized mite extract. In their study, SCIT significantly reduced the total SCORAD in a subgroup of patients with severe AD. One limitation of SCIT is the risk of potential side effects, which include systemic allergic reactions, occasional anaphylaxis, and even fatalities. As serious side effects of SCIT typically occur within 30 minutes of subcutaneous injection, it is necessary to closely monitor the patient for 30 minutes after injection. In one RCT, SLIT with a standardized mite extract was shown to be effective in treating children with mild-to-moderate AD. However, the benefit was inconsistent in the severe form of AD45. There have been no well-organized comparison studies comparing SLIT and SCIT. SLIT is self-administered by patients or their caregivers at home, although the initial dose is usually given under medical supervision. The main advantages of SLIT over SCIT are safety and the convenience of self-administration. Recommendations for ASIT are summarized in Table 953839.

Phototherapy is a common treatment modality in AD patients. Various types of photo (chemo) therapy, light sources, and laser devices can be applied in AD treatment. These include narrowband ultraviolet B (NB-UVB), ultraviolet A1 (UVA1), and light-emitting diodes46. Phototherapy with medium-dose UVA1 can be used to control acute flares of AD. NB-UVB can be applied to manage the chronic stage of AD47. A recent study indicated a beneficial effect of NB-UVB on immune and barrier abnormalities in AD patients. Twelve patients with moderate-to-severe chronic AD received NB-UVB phototherapy three times weekly for up to 12 weeks. All patients achieved a reduction of at least 50% in SCORAD index scores with NB-UVB phototherapy. Moreover, the Th2 and Th1 immune pathways were suppressed and measures of epidermal hyperplasia and differentiation normalized48. There is still no standard protocol for the optimal dose, duration, and frequency of NB-UVB treatment49. The optimal treatment dose of UVA1 has also not yet been determined. Several studies reported that high and medium doses of UVA1 were superior to a low-dose regimen50. Tzaneva et al.51 also reported that medium-dose UVA1 was as effective as high-dose treatment. The comparative efficacy of the two UV treatments was studied in 28 AD patients who received a six-week course of medium-dose UVA1 or NB-UVB52. In addition, the efficacy and tolerability of both modalities may be considered similarly favorable. When the side effects of NB-UVB and UVA1 are compared, the total amount of irradiation necessary for effective phototherapy is lower for NB-UVB than for medium-dose UVA1. The exposure time is therefore shorter and less heat is produced during NB-UVB treatment compared with medium-dose UVA1. NB-UVB therapy is more comfortable, particularly for AD patients, in which heat can be a trigger for itchiness53. As the long-term effects of phototherapy have not been described, treatment should be reserved for adults and children older than 12 years of age with severe, recalcitrant AD23. Considering the low accessibility of UVA1 devices compared to other modalities of phototherapy, NB-UVB offer the most efficacious and cost-effective evidence-based treatment for patients with chronic AD. Recommendations for phototherapy are summarized in Table 10464753.

IFN-γ can be used in patients with severe AD as a second-line systemic immunomodulator treatment after cyclosporine54. However, it is not commonly used in Korea, since the therapeutic effects of IFN-γ in AD patients have not yet been elucidated and some reports showed only a moderate therapeutic effect.

There is no consensus regarding the dosage or administration method. Some reports used a dosage of 50 µg/m² body surface area via subcutaneous injection, either daily or three times a week, depending on the side effects (Table 11)35455. In terms of side effects, 30%~60% of patients experience intermittent headaches, myalgia, and chills, but these can be controlled with acetaminophen and are relatively safe. The use of IFN-γ in children is not recommended because of the lack of clinical data.

Alitretinoin, also known as 9-cis-retinoic acid, is a recently developed retinoid derivative. It can be administered orally in patients with severe AD-related chronic hand eczema (1a, A)56. It is not recommended as a routine treatment because of the lack of evidence from treating areas other than the hands of AD patients.

Thymopentin (Timunox; Janssen, DE) is a synthetic peptide of amino acids 32~36 (Arg-Lys-Asp-Val-Tyr) of the 52-amino acid thymic hormone thymopoietin. Because of the immunoregulatory effects of thymopentin, subcutaneous thymopentin has been tried in patients with AD57. A randomized sampling study of 39 patients, including children and adults with moderate AD, reported that the group treated with thymopentin for 12 weeks showed better average improvement than the placebo group, without specific adverse events57. However, thymopentin is not yet recommended for treating severe AD because of the lack of evidence and follow-up studies19. Some reports have shown intravenous immunoglobulin to be effective in patients with severe AD. Nevertheless, these therapies are not currently recommended for AD treatment, as no significant effects have been demonstrated in an RCT comparing them with cyclosporine and placebo.

Compared to psoriasis, AD appears to be an orphan disease in the development of targeted therapies and biologics. In AD, there are too many potential targets for directed therapeutic attack, impeding the development of novel biologics58. The results of several small trials of biologics, including omalizumab, rituximab, alefacept, and mepolizumab, in AD were intriguing59606162. However, it is clearly premature to recommend off-label use of these biologics for recalcitrant AD, unless other therapies have failed or are contraindicated (Table 12)558. Ongoing studies aim to identify appropriate therapeutic targets.

The use of complementary and alternative therapy in AD is common in Korea. There are insufficient RCT data concerning complementary and alternative therapies, such as traditional Korean medicine and acupuncture. The results generally suggest a limited role and a potential for serious side effects, such as liver toxicity, for complementary therapy in AD treatment. The consensus recommendations of experts also showed the same results (Table 13)717273.