Atopic dermatitis and skin disease

Efficacy and safety of systemic treatments for

moderate-to-severe atopic dermatitis: A systematic review
Evelien Roekevisch, MD,a Phyllis Ira Spuls, MD, PhD,a Denise Kuester, BA,b,c Jacqueline Limpens, PhD,a and
Jochen Schmitt, MD, MPHb,c Amsterdam, The Netherlands, and Dresden, Germany

Background: Many patients with moderate-to-severe atopic trials evaluating long-term treatments are required. (J Allergy
dermatitis (AD) require systemic immunomodulating treatment Clin Immunol 2014;133:429-38.)
to achieve adequate disease control.
Objective: We sought to systematically evaluate the efficacy and Key words: Atopic dermatitis, evidence-based medicine/systematic
safety of systemic treatments for moderate-to-severe AD. review, immunosuppressive therapy, immunomodulator, systemic
Methods: A systematic literature search was performed in treatment, recommendations, GRADE
MEDLINE, EMBASE, and CENTRAL (until June 2012).
Randomized controlled trials (RCTs) evaluating systemic In many patients with moderate-to-severe atopic dermatitis
immunomodulating treatments for moderate-to-severe AD were (AD), disease activity requires systemic treatment to achieve
included. Selection, data extraction, quality assessment, and adequate disease control. Data from routine clinical care
generation of treatment recommendations using the Grading of suggest that more than 10% of all patients with AD receive
Recommendations Assessment, Development and Evaluation systemic anti-inflammatory treatment.1 Various immunomodulat-
(GRADE) approach were performed independently by 2 ing therapies are currently used, including glucocorticosteroids,
reviewers. Efficacy outcomes were clinical signs, symptoms, cyclosporin A (CsA), methotrexate (MTX), azathioprine
quality of life, and the course of AD. Safety data were compared (AZA), IFN-g, intravenous immunoglobulin (IVIG), montelukast,
by calculating the weekly incidence rates (as percentages) for Enteric-coated mycophenolate sodium (EC-MPS), and traditional
adverse events. Chinese herbal medicine (TCHM). We conducted a systematic
Results: Thirty-four RCTs with 12 different systemic treatments review to critically appraise the evidence on the benefits and
and totaling 1653 patients were included. Fourteen trials harms of systemic treatments for AD, to provide clinically relevant
consistently indicate that cyclosporin A efficaciously improves recommendations, and to inform future guideline development.
clinical signs of AD. Cyclosporin A is recommended as first-line
treatment for short-term use. A second-line treatment option is
azathioprine, but efficacy is lower, and evidence is weaker. METHODS
We conducted a systematic review on the efficacy and safety of immuno-
Methotrexate can be considered a third-line treatment option.
modulating systemic treatments (further noted as systemics) for moderate-to-
Recommendations are impossible for mycophenolate, severe AD.
montelukast, intravenous immunoglobulins, and systemic
glucocorticosteroids because of limited evidence. A meta-
analysis was not performed because of a lack of standardization Eligibility criteria
in outcome measures. All fully published randomized controlled trials (RCTs) or open-label
Conclusion: Although 12 different interventions for moderate- extensions of RCTs reporting on systemics in patients with moderate-to-severe
AD were included. Because of the absence of an established definition of
to-severe AD have been studied in 34 RCTs, strong
moderate-to-severe AD, RCTs were eligible when including subjects defined
recommendations are only possible for the short-term use of as patients with moderate-to-severe AD, patients with non-adequately
cyclosporin A. Methodological limitations in the majority of controlled AD despite the use of topical anti-inflammatory therapy, or
trials prevent evidence-based conclusions. Large head-to-head patients with moderate-to-severe AD according to severity criteria (Rajka and
Langeland score2 >4.52, SCORAD score3 >20%, or BSA4 >10%).
RCTs without original data (ie, reviews) or a full report available (eg, letters
From athe Department of Dermatology, Academic Medical Center, Amsterdam; bthe
Centre for Evidence-based Healthcare, University Hospital Dresden; and cthe Institute and abstracts) in which a mixed population of different dermatologic
for Occupational and Social Medicine, Technical University Dresden. conditions was studied without AD being analyzed separately, including
Disclosure of potential conflict of interest: P. I. Spuls is the coauthor of the MACAD trial subgroups of AD, and RCTs not reporting a clinical outcome (clinical signs) or
included in this review, has consultant arrangements for Novartis, and is the principal patient-reported outcome (eg, quality of life and symptoms) were excluded.
investigator of the Department for Clinical Trials with Pharma. J. Schmitt has There were no language restrictions.
consultant arrangements with Novartis and has grants/grants pending with Novartis,
Abbott, MSD, and Wyeth. The rest of the authors declare that they have no relevant
conflicts of interest. Literature search
Received for publication March 25, 2013; revised July 2, 2013; accepted for publication A medical librarian (J.L.) searched MEDLINE (OVID, from 1948),
July 31, 2013. EMBASE (OVID, from 1980), and the Cochrane Central Register of
Available online October 24, 2013.
Controlled Trials (CENTRAL, from inception) to June 2012. The search
Corresponding author: Jochen Schmitt, MD, MPH, Centre for Evidence-based Health-
strategy consisted of (1) free-text words and subject headings related to AD,
care, University Hospital Dresden, Fetscherstrae 74, 01307 Dresden, Germany.
E-mail: [email protected]. (2) specific immunomodulatory drugs or systemic therapy, and (3) a filter to
0091-6749/$36.00 find RCTs.5 The search strategy for MEDLINE is shown in Table E1 in this
2013 American Academy of Allergy, Asthma & Immunology articles Online Repository at www.jacionline.org. The search included an it-
http://dx.doi.org/10.1016/j.jaci.2013.07.049 erative process to refine the search strategy through adding search terms as

429
430 ROEKEVISCH ET AL J ALLERGY CLIN IMMUNOL
FEBRUARY 2014

Strength of evidence and recommendations

Abbreviations used Two reviewers (E.R. and J.S.) graded the quality of evidence and the
AD: Atopic dermatitis strength of the recommendation using the Grading of Recommendations
AE: Adverse event Assessment, Development and Evaluation (GRADE)10 methodology. How-
AZA: Azathioprine ever, in this review GRADE was used in an adjusted way. The evidence per
CsA: Cyclosporin A treatment was graded instead of the evidence per outcome. GRADE could
EASI: Eczema Area and Severity Index only be applied for the effect of treatments on clinical signs because of poor
EC-MPS: Enteric-coated mycophenolate sodium reporting of results in other outcome domains. Discrepancies were resolved
GRADE: Grading of Recommendations Assessment, Development with a third reviewer (P.I.S.). The definitions of high, moderate, low,
and Evaluation and very low were used in grading the quality of evidence, according to
IVIG: Intravenous immunoglobulin the GRADE methodology (Table I and see Appendix E1 in this articles
MTX: Methotrexate Online Repository at www.jacionline.org).10
RCT: Randomized controlled trial
SAE: Serious adverse event
TCHM: Traditional Chinese herbal medicine RESULTS
TP-5: Thymopentin Results of the literature search
Of the 925 references retrieved by using the systematic search,
34 RCTs11-44 were selected for review, including a total of 1653
new relevant citations were identified. Reference Manager software (version patients (see Fig E1 in this articles Online Repository at www.
12.0) was used to manage and deduplicate all identified references. jacionline.org).

Study selection and data extraction

At least 2 reviewers (E.R., D.K., or J.S.) independently conducted the Study characteristics
selection and data extraction. Any discrepancies were resolved with a fourth Twelve different types of systemic treatments for moderate-to-
reviewer (P.I.S.). Author, country, year of publication, number of patients, age severe AD were investigated.
range, sex, severity of AD, study design, dosage and duration of treatment, Fourteen RCTs evaluated CsA.* Five RCTs were placebo
concurrent treatment, clinical outcome measures, and results were extracted. controlled27,32,37,39,41; six were head-to-head comparisons with
For chronic diseases, such as AD, it is necessary to distinguish between short- IVIG,11 UV,17 EC-MPS,18 tacrolimus,28 and prednisolone33; and
term (ie, to induce remission) and long-term treatment. For this review, short- 1 compared a new (Sandimmune) versus an old CsA formulation
term treatment is defined as active treatment of less than 16 weeks duration. (Neoral).44 Three RCTs were dose finding.16,20,43
Three RCTs12,26,34 evaluated AZA, including 82 patients treated
Quality assessment with AZA. Two were placebo-controlled trials,12,26 and 1 RCT34
The Risk of Bias tool was used for quality assessment with review Manager
was a head-to-head RCT versus MTX (n 5 20).
5.1 (Cochrane Collaboration) software to conduct the analyses.6 Extra clarify-
IFN-g was evaluated in 2 RCTs.19,23
ing rules were defined as follows. If the observer was blinded but the patient
and RCT personnel were not blinded, we judged the risk of bias as unclear Three RCTs21,25,33 evaluated systemic glucocorticosteroids,
because of insufficient blinding. If an RCT had dropouts and an intention-to- including 20 children treated with flunisolide25 and an unreported
treat analysis was used but no information on the handling of the missing data number of children with beclomethasone diproprionate.21 Twenty-
was given, the risk of bias was scored unclear because of incomplete one adults were treated with prednisolone.33 Three RCTs11,24,29
outcomes. We made a separate column for authors and sponsors conflict evaluated IVIG, including 45 patients treated with IVIG.
of interest as a potential source for bias.7 When no information on funding Montelukast was investigated in 4 RCTs.14,30,31,40 Two
or conflict of interest was provided, we assumed that there was none. RCTs30,40 were placebo-controlled trials, and 2 RCTs14,31 were
head-to-head trials against antihistamine, a topical steroid/
Data synthesis antibiotic.
In accordance with the Harmonising Outcome Measures for Eczema core Four RCTs15,22,35,36 evaluated TCHM. Mycobacterium
outcome domains for AD trials,8 outcomes concerning clinical signs,
vaccae,13 pimecrolimus,42 and thymopentin (TP-5)38 were
symptoms, health-related quality of life, and course of AD were abstracted
investigated in 1 RCT each, respectively.
as efficacy outcomes. Additionally, the primary outcome measures of included
RCTs and changes in IgE levels and eosinophil counts were extracted. If the Twenty-two trials were conducted in Europe, 2 in the United
primary outcome measure was not explicitly defined, the first outcome States,19,38 and 7 in Asia,11,15,22-24,30,31 and 3 RCTs were interna-
mentioned in the results section was assumed to be the primary outcome. tional.13,17,24 The number of patients ranged from 10 to 166. Most
The mean change in efficacy outcome was defined as investigator-rated RCTs (94%) were small and recruited less than 100 patients.
measurement9 from baseline to the end of active treatment. If not mentioned in Long-term treatment of 16 weeks or greater was assessed in 5
the article, the mean change in clinical severity was calculated by using (15%) studies.
absolute scores at baseline and the end of active treatment extracted from Topical corticosteroids as concomitant therapy were allowed in
the text, a presented figure, or graph. In crossover RCTs only the period before 24 (71%) RCTs.
crossover was used to avoid information bias because of carryover effects.
Ninety-seven percent of the RCTs appeared to have the domain
To compare safety data, the incidence rates (as percentages) per patient
clinical signs as the primary outcome domain, and 1 (3%)
per week for adverse events (AEs), serious adverse events (SAEs), and
withdrawals because of AEs or SAEs were calculated as follows: RCT32 used quality of life as the primary outcome domain. The
34 selected RCTs applied a total of 12 different clinical sign
Number of events=Number of patients 3 Duration of RCT in weeks 3 100:

We intended to pool the mean change in clinical severity of AD from *References 11, 16-18, 20, 27, 28, 32, 33, 37, 39, 41, 43, and 44.
baseline until the end of active treatment of qualitatively homogeneous RCTs References 12, 14, 16, 18, 20, 21, 25-29, 32-37, 39-41, 43, and 44.
in a random-effects meta-analysis. References 11-13, 15, 17-22, 26, 27, 29, 30, 32-35, 37, 38, and 41-44.
J ALLERGY CLIN IMMUNOL ROEKEVISCH ET AL 431
VOLUME 133, NUMBER 2

scales as the primary outcome measure. The (objective) (2.5-3 mg/kg per day) in the short-term treatment of AD.43 Long-
SCORAD score was most frequently used (11 [32%] RCTs), term use of CsA up to 1 year can be recommended based on 4
and the Eczema Area and Severity Index (EASI) was used in 2 trials,17,18,20,43 but evidence is limited because of the open-label
RCTs.40,42 In 17 (50%) RCTs** no validated clinical sign scores design17,20,43 and high dropout rates.17,43
were used. For further details on study characteristics, see Table Harm. The weekly rate of any AEs ranged between 1.0%41
E2 in this articles Online Repository at www.jacionline.org. and 28.2%.33 The weekly rate of withdrawals because of AEs
Twelve RCTs mentioned changes in IgE levels, eosinophil ranged between 0% and 2.0%.33 The weekly rate of any
counts, or both after treatment. One RCT19 showed a significant SAE ranged from 0.0% to 2.2%.39 In 9 RCTs no severe
reduction in eosinophil counts and a possible trend in IgE levels SAEs were reported.***** Unclear information on the
after treatment with IFN-g. A decrease in total serum IgE levels occurrence of SAEs per group was provided in 2 RCTs.11,20
was seen in patients treated with EC-MPS,18 and the difference Recommendation. CsA is currently recommended as the
in IgE levels between those receiving EC-MPS and those first-line short-term treatment option for moderate-to-severe AD
receiving CsA treatment was significant. because of moderate- and high-quality studies based on the
Most RCTs (44%) included both children and adults, GRADE approach and the efficacy and safety shown for
511,13,20,25,30 (15%) included exclusively children, and 14 short-term use, including large patient numbers.
(41%) RCTs included only adults (age > _ 18 years). Details on AZA. Benefit. AZA, a purine analog, was superior to placebo,
study characteristics, study eligibility criteria, and baseline with a mean improvement in Six Area, Six Sign Atopic Dermatitis
characteristics are summarized in Tables E2 and E3 in this scores of 26%12 and 37%26 at week 12. AZA and MTX were
articles Online Repository at www.jacionline.org. found to be equally efficacious, with a mean SCORAD score
improvement of approximately 39%34 for AZA and a mean
improvement of 20% in quality of life (SKINDEX) after 12
Study quality/risk of bias weeks. Sixteen (84%) patients in the AZA group continued. At
Authors often reported that the trial was randomized without week 24, the relative reduction in the mean SCORAD score
stating how the randomization sequence was generated. An was 43%.
unclear or high risk of bias was found in 60%*** (21/34 Harm. The weekly rate of any AE ranged between 5.6%26 and
RCTs) for sequence generation, 79% (27/34 RCTs) for 22.9%.34 Abnormalities in blood counts, such as lymphocytope-
allocation concealment, 63% (22/34 RCTs) for blinding, nia, were most frequently seen. The weekly rate of withdrawals
71% (24/34 RCTs) for incomplete outcome data, 24%**** because of AEs ranged between 0.2%8 and 0.4%.26 SAEs were
(8/34 RCTs) for selective outcome reporting, and 32% not observed12,34 or not reported.26 After 24 weeks, no
(11/34 RCTs) for other biases (see Figs E2 and E3 in this articles SAEs were seen, in 9% AEs led to withdrawal, and in 9% dose
Online Repository at www.jacionline.org). adjustments were related to AEs. On the basis of the RCT
by Meggitt,26 the AZA dosage should be determined on
the basis of thiopurine S-methyltransferase activity to limit
Efficacy and safely of systemic treatments for myelosuppression.
moderate-to-severe AD Recommendation. AZA is currently recommended as a
An overview of the (modified) efficacy results can be found in second-line treatment option for moderate-to-severe AD because
Table II. The rest of the efficacy and safety results can be found in of a moderate-quality study based on the GRADE approach and
Tables E4 and E5 in this articles Online Repository at www. the efficacy and safety shown for short- and long-term use
jacionline.org. Recommendations according to the GRADE (24 weeks), including large patient numbers.
approach are provided in Table I. MTX. Benefit. MTX, a folic acid antagonist, was equally
CsA. Benefit. CsA, which inhibits the proliferation of T efficacious as AZA, with a mean improvement in SCORAD score
lymphocytes, was superior to placebo in 5 RCTs, with a mean of 42% and a mean improvement of 26% in quality of life
clinical improvement in severity between 53% and 95% in (SKINDEX) after 12 weeks of treatment in a head-to-head trial.34
different clinical severity scores after short-term treatment Eighteen (95%) patients in the MTX group continued. At
(10 days to 8 weeks).27,37,39,41 One RCT used quality of life as week 24, the relative reduction in the mean SCORAD score
the primary outcome and found CsA to be superior to placebo.32 was 48%.
In head-to-head trials CsA was superior to prednisolone,33 supe- Harm. The weekly rate of any AE was 23.5%. The weekly
rior to IVIG,11 superior to UVAB,17 and similarly efficacious as rate of withdrawals because of AEs was 0.4%. SAEs were
EC-MPS.18 Higher CsA dosages (5 mg/kg per day) lead to a not observed. After 24 weeks, no SAEs were seen, in 5% AEs
more rapid response and are more efficacious than lower dosages led to withdrawal, and in 10% dose adjustments were related to
AEs.
Recommendation. MTX is recommended as a third-line
References 11, 14, 17, 18, 22, 24, 28, 29, 31, 33, and 34. treatment for adults with severe AD because of a moderate-
**References 15, 16, 19, 21-23, 25, 27, 30, 35-39, 41, 43, and 44.
quality study based on the GRADE approach and the efficacy and
References 14, 15, 18, 19, 21-24, 28, 29, 36, and 37.
References 12, 15, 19, 22-24, 26, 28, 31, 32, 35, 36, and 38-40. safety shown for short- and long-term use (24 weeks), including
References 14, 16-18, 21, 27, 29, 33, 34, 37, and 41-44. large patient numbers.
***References 11, 14, 17, 19, 20, 23-25, 27, 28, 31, 32, and 35-43.
References 11, 14-21, 23-25, 27-29, 31, 32, 34-41, 43, and 44.
References 11, 13, 14, 17-20, 23-25, 29, 31-34, 36, and 39-44.
References 11-14, 17, 19-21, 23-25, 27, 29, 30, 32, 35-41, 43, and 44. References 16, 20, 27, 28, 32, 37, 41, and 44.
****References 11, 12, 14, 18, 23, 26, 29, and 32. References 16-18, 27, 28, 33, 41, 43, and 44.
References 11, 12, 14-16, 18, 21, 23, 33, 35, and 36. *****References 16-18, 27, 28, 33, 41, 43, and 44.
432 ROEKEVISCH ET AL J ALLERGY CLIN IMMUNOL
FEBRUARY 2014

TABLE I. GRADE
No. of RCTs (total participants) Comparison Risk of bias* Inconsistency* Indirectness* Imprecision* Publication bias*

5 (146) 27,32,37,39,41
CsA vs placebo 0 0 0 0 21

2 (182)16,43 CsA, higher dose vs lower dose 21 0 0 0 0

1 (14)44 CsA, older vs newer formulation 0 0 0 22 21

1 (38)35 CsA vs prednisolone 0 0 0 21 21

1 (50)18 CsA vs EC-MPS 22 0 0 0 21

1 (14)11 CsA vs IVIG 21 0 0 22 0

1 (30)28 CsA vs topical tacrolimus 21 0 0 21 0

1 (72)17 CsA vs UVAB 21 0 0 21 21

2 (98)12,26 AZA vs placebo 0 21 0 21 21

1 (42)34 AZA vs MTX 0 0 0 21 0

1 (103)42 Pimecrolimus vs placebo 0 0 0 21 21

2 (134)19,23 IFN-g vs placebo 21 0 0 0 21

2 (67)21,25 Systemic glucocorticosteroids vs 21 0 0 21 21

placebo
2 (74)30,40 Montelukast vs placebo 0 22 0 21 0

2 (63)14,31 Montelukast vs antihistamine, 21 22 0 21 0

topical steroid 6 antibiotic

1 (166)13 M vaccae vs placebo 21 0 0 0 0

1 (39)38 TP-5 vs placebo 21 0 0 21 0

4 (79)15,22,35,36 TCHM vs placebo 21 22 0 0 0

2 (49)24,29 IVIG vs placebo/waiting control 22 0 0 21 0

*Downgrading rules:
Risk of Bias (consideration of items 1-6 [without conflict of interest]): No item high in majority of studies / no downgrading/1 item high in majority of studies / 21/2 or more
items high in majority of studies / 22.
Inconsistency of the results: Maximal downgrading: 22/downgrading, if substantial variation in effect between studies, only quantitative: 21, also qualitative: 22
Indirectness of the evidence: Maximal downgrading: 21/patient population highly selected (eg, only subgroup of patients with AEs): 21
Imprecision of the results: Maximal downgrading: 22/if sum of patients for 1 comparison in all trials is <100: / downgrade 21/if sum of patients for 1 comparison in all trials
<20 / downgrade 22.
Publication bias: Maximal downgrading: 22.
Upgrading rules:
Large overall effect: Does not apply to RCTs
Dose-response relationship: If dose-response relationship is present: 11
Confounder: Does not apply to RCTs
XFS, Xiao-Feng-San.

IFN-g. Benefit. Both RCTs19,23 investigating IFN-g, which Harm. The weekly rate of any AE was 16.3% in one trial19 and
downregulates TH2 cell function, versus placebo found IFN-g to be not reported in the other trial.23
efficacious after 12 weeks of treatment based on unvalidated outcome Recommendation. The quality of evidence of efficacy
measures. One trial23 observed a positive dose-response relationship. of IFN-g versus placebo is moderate for short-term use. IFN-g
J ALLERGY CLIN IMMUNOL ROEKEVISCH ET AL 433
VOLUME 133, NUMBER 2

TABLE I. (Continued)
Dose responsey GRADE Conclusion

0 Moderate quality CsA is effective in patients with moderate-to-severe AD. Further research is likely to have an important effect on our
confidence in the estimate of effect and might change our recommendation.
1 High quality Higher dosages of CsA are more effective than lower dosages of CsA in patients with severe AD. Further research is
very unlikely to change our confidence in the estimate of effect.
0 Very low quality Current evidence on the comparative efficacy of older formulation CsA versus newer formulation CsA for AD is very
weak and suggests equal effectiveness of both CsA formulations. Any estimate of effect in future research is very
uncertain.
0 Low quality CsA is more effective than prednisolone to induce stable remission in patients with severe AD. Further research is very
likely to have an important effect on our confidence in the estimate of effect and is likely to change our
recommendation.
0 Very low quality CsA and EC-MPS might be equally effective as a maintenance treatment in some patients with severe AD, but our
confidence in the effect is very low. Any estimate of effect in future research is very uncertain.
0 Very low quality Current evidence is insufficient to recommend the use of IVIG for severe AD. Current evidence on the comparative
effectiveness of CsA versus IVIG for AD is very weak and favors CsA. Any estimate of effect in future research is
very uncertain.
0 Low quality CsA and topical tacrolimus 0.1% might be equally effective in patients with moderate-to-severe AD, but our confidence
in the effect is low. Further research is very likely to have an important effect on our confidence in the estimate of
effect and is likely to change our recommendation.
0 Very low quality CsA might be more effective than UVAB in patients with severe AD, but our confidence in the effect is low. Any
estimate of effect in future research is very uncertain.
0 Very low quality AZA is effective in many patients with moderate-to-severe AD. Any estimate of effect in future research is very
uncertain.
0 Moderate quality MTX and AZA are both equally effective in patients with severe AD and might be considered as a treatment option.
Further research is likely to have an important effect on our confidence in the estimate of effect and might change
our recommendation.
1 Moderate quality Systemic pimecrolimus is effective in some patients with moderate-to-severe AD and can be considered as a treatment
option. Further research is likely to have an important effect on our confidence in the estimate of effect and might
change our recommendation.
1 Moderate quality IFN-g is effective in some patients with severe AD and can be considered as a treatment option. Further research is
likely to have an important effect on our confidence in the estimate of effect and might change our recommendation.
0 Very low quality Systemic glucocorticosteroids (beclomethasone diproprionate and flunisolide) might be effective to induce remission in
patients with moderate-to-severe AD. Any estimate of effect in future research is very uncertain.
0 Very low quality We do not know whether montelukast is effective in a subgroup of patients with moderate-to-severe AD. Any estimate
of effect in future research is very uncertain.
0 Very low quality We do not know whether montelukast is more, equally, or less effective as combined treatment with antihistamine,
topical steroid 6 antibiotic in patients with moderate-to-severe AD. Any estimate of effect in future research is
very uncertain.
0 Moderate quality M vaccae is not effective for patients with severe AD and is not recommended based on current evidence. Further
research is likely to have an important effect on our confidence in the estimate of effect and might change our
recommendation.
0 Low quality Current evidence does not recommend the use of TP-5 for patients with severe AD. Further research is very likely to
have an important effect on our confidence in the estimate of effect and is likely to change our recommendation.
0 Very low quality Current evidence is insufficient to recommend the use of XFS or TCHM for patients with moderate-to-severe AD.
Any estimate of effect in future research is very uncertain.
0 Very low quality Current evidence is insufficient to recommend the use of IVIG for patients with severe AD. Any estimate of effect in
future research is very uncertain.

can be considered a third-line treatment option for patients with Within 30 weeks of maintenance treatment, EC-MPS and CsA
severe AD, but safety and tolerability need to be monitored closely. were equally effective.
EC-MPS. Benefit. One small trial18 (n 5 50) investigated the Harm. The weekly rate of any AE was 4.5%. The weekly rate
efficacy of EC-MPS, a purine biosynthesis inhibitor, versus CsA of withdrawals because of AEs was 0.3%. SAEs were not
as a maintenance treatment after induction treatment with CsA. observed.
434 ROEKEVISCH ET AL J ALLERGY CLIN IMMUNOL
FEBRUARY 2014

TABLE II. Summary of efficacy of systemic treatments for moderate-to-severe AD in the RCTs included
Efficacy drug A Efficacy drug B Efficacy placebo
a) Intervention a) Intervention a) Placebo
Reference b) Change in clinical signs b) Change in clinical signs b) Change in clinical signs

Berth-Jones12 a) AZA a) Placebo

b) MI of 26% in SASSAD score at 8 wk b) MI 3% in SASSAD score at 8 wk
Meggit26 a) AZA a) Placebo
b) MI in of 37% in SASSAD score at 12 wk b) MI of 20% in SASSAD score at 12 wk
Bemanian11 a) CsA a) IVIG
b) MI of 68% in SCORAD at 12 wk b) MI of 30% in SCORAD at 12 wk
Czech16 a) CsA, 150 mg a) CsA, 300 mg
b) MI of 48% in TBSA at 8 wk b) MI of 59% in TBSA at 8 wk
Granlund17 a) CsA a) UVAB
b) MI of 54% in SCORAD at 8 wk b) MI of 34% in SCORAD at 8 wk
Haeck18 a) CsA a) EC-MPS
b) MI of 17% in SCORAD score at MI of 0% in SCORAD at the maintenance
maintenance phase 0-30 wk and 42% MI phase 0-30 wk and 39% MI SCORAD
in SCORAD score including run-in phase including run-in phase of 6 wk (26 wk to
of 6 wk (26 wk to 30 wk) 30 wk)
Harper20 a) CsA, short course a) CsA, continuous
b) MI of 42% in SASSAD score at 52 wk b) MI of 56% in SASSAD score at 52 wk
Munro27 a) CsA a) Placebo
b) MI of 95% in score at 8 wk b) MI of 19% in score at 8 wk
Pacor28 a) CsA a) Tacrolimus
b) MI of 88% in area under the SCORAD b) MI of 89% in area under the SCORAD
curve at 6 wk curve at 6 wk
Salek32 a) CsA a) Placebo
b) NR (disease activity and extent of disease) b) NR (disease activity and extent of disease)
Schmitt33 a) CsA a) Prednisolone/placebo
b) MI of 55% in objective SCORAD50 at b) MI of 43% in objective SCORAD50 at
6 wk 6 wk
Sowden37 a) CsA a) Placebo
b) MI of 53% in score at 8 wk b) MI of 7% in score at 8 wk
van Joost39 a) CsA a) Placebo
b) MI of 56% in TBSA at 6 wk b) MI of 11% in TBSA at 6 wk
Wahlgren41 a) CsA a) Placebo
b) MI of 58% in score at 10 d b) Mean deterioration of 4% in score at
10 d
Zonneveld43 a) CsA increasing dose a) CsA, decreasing dose
b) MI of 46% in score at 52 wk b) MI of 29% in score at 52 wk
Zurbriggen44 a) Sandimmune (old formulation of CsA) a) Neoral (new formulation of CsA)
b) MI of 73% in score at 8 wk b) MI of 72% in score at 8 wk
la Rosa25 a) Flunisolide a) Placebo
b) MI of 54%* in TCS at 2 wk b) MI of 16%* in TCS at 2 wk
Hanifin19 a) IFN-g a) Placebo
b) Improvement in TCS is greater than in b) NR
placebo group but not significant at 12 wk
Jang23 a) IFN-g, 1.5 3 106 IU/m2 a) IFN-g, 0.5 3 106 IU/m2 a) Placebo
b) MI 48%* in TCS at 12 wk b) MI 36%* in TCS at 12 wk b) MI 2%* in TCS at 12 wk
Jee24 a) IVIG a) Placebo
b) MI of 24% in SCORAD score at 12 wk b) MI of 40% in SCORAD score at 12 wk
Paul29 a) IVIG a) IVIG delayed
b) MI of 15% in SCORAD score at 4 wk b) MI of 15% in SCORAD score at 4 wk
Berth-Jones12 a) M vaccae, low dose a) M vaccae, high dose a) Placebo
b) MI of 25% in SASSAD score at 12 wk b) MI of 26% in SASSAD score at 12 wk b) MI of 24% in SASSAD score at 12 wk
Schram34 a) MTX/folate a) AZA
b) MI of 42% in SCORAD score at 12 wk b) MI of 39% in SCORAD score at 12 wk
Capella14 a) Montelukast/placebo/nongreasy gel a) Cetirizine/clarithromycin/mometasone or
methylprednisolone/hydrating topical
preparations
b) MI of 51% in SCORAD score at 6 wk b) MI of 50% in SCORAD score at 6 wk
Rahmen31 a) Montelukast a) Antihistamines/hydrocortisone
b) MI of 29% in SCORAD score at 4 wk b) MI of 9% in SCORAD score at 4 wk
(Continued)
J ALLERGY CLIN IMMUNOL ROEKEVISCH ET AL 435
VOLUME 133, NUMBER 2

TABLE II. (Continued)

Efficacy drug A Efficacy drug B Efficacy placebo
a) Intervention a) Intervention a) Placebo
Reference b) Change in clinical signs b) Change in clinical signs b) Change in clinical signs

Veien40 a) Montelukast a) Placebo

b) MI of 24% in validated EASI at 4 wk b) MI of 20% in validated EASI at 4 wk
Pei30 a) Montelukast a) Placebo
b) MI of 42% in score at 4 wk b) MI of 33% in score at 4 wk
Heddle21 a) Oral 1 nasal beclomethasone a) Rotacaps 1 nasal aerosol
diproprionate
b) significantly greater improvements on b) NR
beclomethasone diproprionate than on
placebo at 4 wk
Wolff42 a) Pimecrolimus, 10 mg a) Pimecrolimus, 20 mg (pimecrolimus, a) Placebo
30 mg)
b) MI of 21% in EASI at 12 wk b) MI of 35% in EASI at 12 wk (MI of 47% b) MI of 20% in EASI at 12 wk
in EASI at 12 wk)
Stiller38 a) TP-5 a) Placebo
b) MI of 21% in total severity score at 12 wk b) MI of 14% in total severity score at 12 wk
Hon22 a) TCHM a) Placebo
b) MI of 15% in SCORAD score at 12 wk b) MI of 19% in SCORAD score at 12 wk
Sheehan35 a) TCHM a) Placebo
b) MI of 82% in score at 8 wk b) MI of 20% in score at 8 wk
Sheehan36 a) TCM a) Placebo
b) MI of 60% in score at 8 wk b) MI of 11% in score at 8 wk
Cheng15 a) XFS a) Placebo
b) MI of 56% in score at 8 wk b) MI of 11% in score at 8 wk
Proportion of persons with 5-point or greater improvement (ie, minimum relevant improvement) in DLQI.
DQLI, Dermatological Quality of Life Index; EDI, eczema disability index; MI, mean improvement; NR, not reported; SASSAD, Six Area, Six Sign Atopic Dermatitis; TBSA, total
body surface area; TCS, total clinical severity; XFS, Xiao-Feng-San.
*Values taken out of figure/graph.
Nonvalidated score.

Recommendation. Only a very weak recommendation is with a mean SCORAD score improvement of 51% at week 6, and
possible for EC-MPS as a maintenance treatment for severe AD it was superior to a combination of an antihistamine and topical
after induction of remission by CsA for long-term use up to glucocorticoid after 4 weeks.31
30 weeks. Harm. None of the trials reported AEs of montelukast.
Systemic glucocorticosteroids. Benefit. Two small trials Recommendation. Because of the contradictory results
investigated the short-term efficacy of the systemic glucocorticoste- from the published trials, montelukast is currently not recom-
roids beclomethasone diproprionate21 and flunisolide25 in children mended for the treatment of moderate-to-severe AD.
with severe AD based on unvalidated outcome measures. Systemic Pimecrolimus. Benefit. Systemic pimecrolimus, a calci-
prednisolone was less efficacious than CsA in adults with severe neurin inhibitor, was superior to placebo in a dose-dependent
AD and induced stable remission in only 1 of 21 patients.33 manner with a mean clinical improvement in EASI of 35%
Harm. The weekly rate of any AE ranged between 0%25 and and 47% after 12 weeks of treatment with 20 and 30 mg of
20.4%.33 The study that compared prednisolone versus CsA was pimecrolimus, respectively.42
stopped prematurely because of high rates of exacerbations and Harm. The weekly rate of any AE was 7.7% in the 30-mg
adverse effects after discontinuation of prednisolone treatment.33 group.
Recommendation. Systemic steroids are currently not Recommendation. Systemic pimecrolimus could be a pos-
recommended for moderate-to-severe AD. sible treatment option for moderate-to-severe AD in short-term use.
IVIG. Benefit. IVIG was less efficacious than placebo and TCHM. Benefit. TCHM was shown to be superior to placebo
CsA.11 in several trials, with a mean improvement in an unvalidated score
Harm. The weekly rate of any AE ranged between 0.6%24 and of 56%,22 60%,36 and 82%35 after 8 weeks of treatment. In 1 RCT
2.8%.11 placebo tended to be more efficacious than TCHM, with a mean
Recommendation. IVIG is currently not recommended for improvement in SCORAD score of 15% (TCHM) versus 19%
the treatment of severe AD. (placebo) after 12 weeks of treatment. For further information
Monteluklast. Benefit. Four RCTs14,30,31,40 investigated on the formulations used in the different trials, see Fig E1.
the efficacy of montelukast for moderate-to-severe AD. The re- Harm. The weekly rate of any AE ranged between 0.5%35 and
sults of these trials are inconsistent, indicating beneficial effects 10.8%.22 The weekly rate of withdrawals because of AEs ranged
compared with placebo based on an unvalidated score30 but no between 0.0%35 and 0.2%.15
difference in efficacy compared with placebo in mean EASI Recommendation. Because of the inconsistent results from
change.40 Montelukast was equally effective as a combination the published trials, TCHM is currently not recommended for the
of cetirizine, clarithromycin, and a topical glucocorticosteroid,14 treatment of moderate-to-severe AD.
436 ROEKEVISCH ET AL J ALLERGY CLIN IMMUNOL
FEBRUARY 2014

M vaccae. Benefit. M vaccae was found not to be more Systemic glucocorticosteroids are the most frequently used
efficacious than placebo in children with severe AD.13 systemic treatment for severe AD in routine care.1 In contrast, trial
Harm. AEs per group were not reported. SAEs occurred with a evidence concerning efficacy is very limited. One small trial33
weekly rate of 0.2%. indicated that systemic prednisolone is not suitable to induce
Recommendation. M vaccae is not recommended for the long-term remission and also less efficacious than CsA. However,
treatment of severe AD. it is unclear whether this RCT investigated the right systemic
TP-5. Benefit. One RCT38 based on 39 children and adults steroid treatment regimen. The suggestions on how to apply
with severe AD reported a mean improvement in an unvalidated systemic steroids differ substantially, with some authors suggest-
clinical score of 21% in the TP-5 group versus 12% in the placebo ing treatment only for a few days and others over a period of
group (P 5 .055). months.
Harm. AEs were not reported. Withdrawals because of AEs or Montelukast and TCHM are currently not recommended for
SAEs were not observed. the treatment of moderate-to-severe AD in routine care because
Recommendation. TP-5 is currently not recommended for the results from the trials published are inconsistent. One
the treatment of severe AD. difficulty in assessing TCHM is its heterogeneity. Standard
regimens include a variety of herbs, which can be called
DISCUSSION polypharmaceuticals, and are used in the form of tea. Potential
concerns are the side effects of TCHM, such as liver failure,
This systematic review provides an evidence-based treat-
cardiomyopathy, and hepatotoxity,51-54 although none of these
ment algorithm for patients with moderate-to-severe AD. By
effects were reported in the included RCTs.
ranking the quality of evidence based on the GRADE
approach45 and by taking into account the efficacy and IVIG, M vaccae, and TP-5 are currently not recommended
because trials do not suggest that these treatments are efficacious
safety shown for short-term (and long-term) use and the
in the treatment of moderate-to-severe AD.
number of participants, this review extends previous research
and guidelines in which less specific treatment recommen-
dations were made based on expert opinion46 or less system-
Limitations of this review and published trials
Although 12 different interventions have been studied in
atic approaches, such as the Goodman method47 or the
Category of Evidence and Strength of Recommendation 34 RCTs, strong recommendations are only possible for the
short-term use of CsA. Methodological limitations, such as small
based approach.48-50
sample size, short duration of most of the trials conducted, a lack
of head-to-head trials on key comparators, and unclear or high
Recommendations for systemic treatment of risk of bias in many important domains in a substantial proportion
moderate-to-severe AD of trials, prevent strong evidence-based recommendations in
Fourteen trials consistently indicate that CsA effica- accordance with the GRADE approach.45
ciously improves clinical signs of AD in children and adults. Meta-analysis was not indicated because of the clinical and
Although higher doses (5 mg/kg body weight) of CsA are methodological heterogeneity, such as the lack of standardization
more effective, lower starting doses (3 mg/kg body weight) in outcome measures, the substantial differences in trial designs,
with stepwise adjustment to the individual minimum effective and the wide use of unvalidated outcome measures. Because of
dose are preferable because most side effects are dose the lack of reporting of quality-of-life outcomes and symptoms,
related.16,43 Overall, CsA is recommended as first-line treat- GRADE could only be applied for clinical efficacy based on
ment for short-term use. Long-term use of CsA up to 1 year measures of clinical signs (measures).
can be recommended based on the results of 4 trials.17,18,20,43 Most systemics included in our review are known to be
However, the long-term safety of CsA cannot be concluded associated with potential adverse drug reactions.47,49,55-57 How-
from the trial evidence. ever, mainly short-term RCTs were included in this review;
As a second-line treatment option, AZA is recommended for 85% (29/34) of the RCTs had a total RCT period of less than
short-term induction treatment and long-term treatment up to 24 16 weeks. Important AEs, SAEs, and withdrawals that occur after
weeks. The dosage should be determined on the basis of long-term treatment or follow-up could have been missed.
thiopurine S-methyltransferase activity to limit myelosuppres- Many RCTs provided inadequate information on AEs. Better
sion.26 Indirect comparisons suggest that the efficacy of AZA is arrangements need to be made on how to document short- and
lower than that of CsA. long-term safety data. Prospective registries with these drugs
MTX can be considered a third-line treatment option for short- could detect rare and long-term side effects.
term induction treatment and long-term treatment up to 24 weeks,
but the evidence is limited to a rather small single trial that found Research recommendations
MTX to be similarly efficacious as AZA.34 More head-to-head trials or prospective registries are required
According to placebo-controlled trials, IFN-g is also efficacious to draw relevant conclusions for routine care and clarify the
for severe AD. IFN-g can be considered a third-line treatment comparative effectiveness, safety, and tolerability of CsA, AZA,
option, but safety and tolerability need to be monitored closely. MTX, IFN-g, systemic steroids, TP-5, and EC-MPS. Such trials
EC-MPS might be a treatment option for maintenance should consider the 4 core outcome domains for clinical trials, as
treatment of AD after induction treatment with CsA, but this defined by the Harmonising Outcome Measures for Eczema
recommendation can currently be based only on a single study.18 initiative.8
Although the prevalence of AD is highest among children,
RCTs on children are missing for many relevant interventions,
References 11, 16-18, 20, 27, 28, 32, 33, 37, 39, 41, 43, and 44. and more research in this age group is very important.
J ALLERGY CLIN IMMUNOL ROEKEVISCH ET AL 437
VOLUME 133, NUMBER 2

dermatitis: a randomized, double-blind, placebo-controlled study. Br J Dermatol

Clinical implications: CsA is recommended as first-line treat- 2007;157:357-63.
ment for short-term use in patients with moderate-to-severe 23. Jang IG, Yang JK, Lee HJ, Yi JY, Kim HO, Kim CW, et al. Clinical improvement
AD. AZA is the second-line and MTX the third-line and immunohistochemical findings in severe atopic dermatitis treated with inter-
recommendation. feron gamma. J Am Acad Dermatol 2000;42:1033-40.
24. Jee SJ, Kim JH, Baek HS, Lee HB, Oh JW. Long-term efficacy of intravenous im-
munoglobulin therapy for moderate to severe childhood atopic dermatitis. Allergy
REFERENCES Asthma Immunol Res 2011;3:89-95.
1. Schmitt J, Schmitt NM, Kirch W, Meurer M. Outpatient care and medical treat- 25. La Rosa M, Musarra I, Ranno C, Matello N, Negri L, Miraglia del Guidice JR M,
ment of children and adults with atopic eczema. J Dtsch Dermatol Ges 2009;7: et al. A randomized, double-blind, placebo-controlled crossover trial of systemic
345-51. flunisolide in the treatment of children with severe atopic dermatitis. Curr Ther
2. Rajka G, Langeland T. Grading of the severity of atopic dermatitis. Acta Derm Res 1995;56:720-6.
Venereol Suppl (Stockh) 1989;144:13-4. 26. Meggitt SJ, Gray JC, Reynolds NJ. Azathioprine dosed by thiopurine methyl-
3. Kunz B, Oranje AP, Labreze L, Stalder JF, Ring J, Taieb A. Clinical validation transferase activity for moderate-to-severe atopic eczema: a double-blind, rando-
and guidelines for the SCORAD index: consensus report of the European Task mised controlled trial. Lancet 2006;367:839-46.
Force on Atopic Dermatitis. Dermatology 1997;195:10-9. 27. Munro CS, Levell NJ, Shuster S, Friedmann PS. Maintenance treatment with
4. Lee SS, Tan AW, Giam YC. Cyclosporin in the treatment of severe atopic derma- cyclosporin in atopic eczema. Br J Dermatol 1994;130:376-80.
titis: a retrospective study. Ann Acad Med Singapore 2004;33:311-3. 28. Pacor ML, Di LG, Martinelli N, Mansueto P, Rini GB, Corrocher R. Comparing
5. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of tacrolimus ointment and oral cyclosporine in adult patients affected by atopic
interventions, version 5.1.0 [updated March 2011]. The Cochrane Collaboration; dermatitis: a randomized study. Clin Exp Allergy 2004;34:639-45.
2011. Available at: www.cochrane-handbook.org. Accessed October 17, 2013. 29. Paul C, Lahfa M, Bachelez H, Chevret S, Dubertret L. A randomized controlled
6. Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. The evaluator-blinded trial of intravenous immunoglobulin in adults with severe
Cochrane Collaborations tool for assessing risk of bias in randomised trials. atopic dermatitis. Br J Dermatol 2002;147:518-22.
BMJ 2011;343:d5928. 30. Pei AY, Chan HH, Leung TF. Montelukast in the treatment of children with
7. Sharek Z, Schoen RE, Loewenstein G. Bias in the evaluation of conflict of inter- moderate-to-severe atopic dermatitis: a pilot study. Pediatr Allergy Immunol 2001;
est policies. J Law Med Ethics 2012;40:368-82. 12:154-8.
8. Schmitt J, Spuls P, Boers M, Thomas K, Chalmers J, Roekevisch E, et al. Towards 31. Rahman ML, Choudhury AM, Islam MM. Effectiveness of montelukast in the
global consensus on outcome measures for atopic eczema research: results of the treatment of atopic dermatitis. Mymensingh Med J 2006;15:85-8.
HOME II meeting. Allergy 2012;67:1111-7. 32. Salek MS, Finlay AY, Luscombe DK, Allen BR, Berth-Jones J, Camp RD, et al.
9. Schmitt J, Langan S, Williams HC. What are the best outcome measurements Cyclosporin greatly improves the quality of life of adults with severe atopic
for atopic eczema? A systematic review. J Allergy Clin Immunol 2007;120: dermatitis. A randomized, double-blind, placebo-controlled trial. Br J Dermatol
1389-98. 1993;129:422-30.
10. Schunemann HJ, Oxman AD, Brozek J, Glasziou P, Bossuyt P, Chang S, et al. 33. Schmitt J, Schakel K, Folster-Holst R, Bauer A, Oertel R, Augustin M, et al.
GRADE: assessing the quality of evidence for diagnostic recommendations. Prednisolone vs. ciclosporin for severe adult eczema. An investigator-initiated
ACP J Club 2008;149:2. double-blind placebo-controlled multicentre trial. Br J Dermatol 2010;162:661-8.
11. Bemanian MH, Movahedi M, Farhoudi A, Gharagozlou M, Seraj MH, Pourpak Z, 34. Schram ME, Roekevisch E, Leeflang MM, Bos JD, Schmitt J, Spuls PI. A ran-
et al. High doses intravenous immunoglobulin versus oral cyclosporine in the domized trial of methotrexate versus azathioprine for severe atopic eczema.
treatment of severe atopic dermatitis. Iran J Allergy Asthma Immunol 2005;4: J Allergy Clin Immunol 2011;128:353-9.
139-43. 35. Sheehan MP, Rustin MH, Atherton DJ, Buckley C, Harris DW, Brostoff J, et al.
12. Berth-Jones J, Takwale A, Tan E, Barclay G, Agarwal S, Ahmed I, et al. Azathi- Efficacy of traditional Chinese herbal therapy in adult atopic dermatitis [erratum
oprine in severe adult atopic dermatitis: a double-blind, placebo-controlled, cross- appears in Lancet 1992;340:188]. Lancet 1992;340:13-7.
over trial. Br J Dermatol 2002;147:324-30. 36. Sheehan MP, Atherton DJ. A controlled trial of traditional Chinese medicinal plants
13. Berth-Jones J, Arkwright PD, Marasovic D, Savani N, Aldridge CR, Leech SN, in widespread non-exudative atopic eczema. Br J Dermatol 1992;126:179-84.
et al. Killed Mycobacterium vaccae suspension in children with moderate-to- 37. Sowden JM, Berth-Jones J, Ross JS, Motley RJ, Marks R, Finlay AY, et al. Dou-
severe atopic dermatitis: a randomized, double-blind, placebo-controlled trial. ble-blind, controlled, crossover study of cyclosporin in adults with severe refrac-
Clin Exp Allergy 2006;36:1115-21. tory atopic dermatitis. Lancet 1991;338:137-40.
14. Capella GL, Grigerio E, Altomare G. A randomized trial of leukotriene receptor 38. Stiller MJ, Shupack JL, Kenny C, Jondreau L, Cohen DE, Soter NA. A double-
antagonist montelukast in moderate-to-severe atopic dermatitis of adults. Eur J blind, placebo-controlled clinical trial to evaluate the safety and efficacy of thy-
Dermatol 2001;11:209-13. mopentin as an adjunctive treatment in atopic dermatitis. J Am Acad Dermatol
15. Cheng HM, Chiang LC, Jan YM, Chen GW, Li TC. The efficacy and safety of a 1994;30:597-602.
Chinese herbal product (Xiao-Feng-San) for the treatment of refractory atopic 39. van Joost T, Heule F, Korstanje M, van den Broek MJ, Stenveld HJ, van Vloten
dermatitis: a randomized, double-blind, placebo-controlled trial. Int Arch Allergy WA. Cyclosporin in atopic dermatitis: a multicentre placebo-controlled study.
Immunol 2011;155:141-8. Br J Dermatol 1994;130:634-40.
16. Czech W, Brautigam M, Weidinger G, Schopf E. A body-weight-independent 40. Veien NK, Busch-Sorensen M, Stausbol-Gron B. Montelukast treatment of mod-
dosing regimen of cyclosporine microemulsion is effective in severe atopic der- erate to severe atopic dermatitis in adults: a randomized, double-blind, placebo-
matitis and improves the quality of life. J Am Acad Dermatol 2000;42:653-9. controlled trial. J Am Acad Dermatol 2005;53:147-9.
17. Granlund H, Erkko P, Remitz A, Langeland T, Helsing P, Nuutinen M, et al. Com- 41. Wahlgren CF, Scheynius A, Hagermark O. Antipruritic effect of oral cyclosporin
parison of cyclosporin and UVAB phototherapy for intermittent one-year treat- A in atopic dermatitis. Acta Derm Venereol 1990;70:323-9.
ment of atopic dermatitis. Acta Derm Venereol 2001;81:22-7. 42. Wolff K, Fleming C, Hanifin J, Papp K, Reitamo S, Rustin M, et al. Efficacy and
18. Haeck IM, Knol MJ, Ten BO, van Velsen SG, de Bruin-Weller MS, Bruijnzeel- tolerability of three different doses of oral pimecrolimus in the treatment of mod-
Koomen CA. Enteric-coated mycophenolate sodium versus cyclosporin A as erate to severe atopic dermatitis: a randomized controlled trial. Br J Dermatol
long-term treatment in adult patients with severe atopic dermatitis: a randomized 2005;152:1296-303.
controlled trial. J Am Acad Dermatol 2011;64:1074-84. 43. Zonneveld IM, De Rie MA, Beljaards RC, Van Der Rhee HJ, Wuite J, Zeegelaar
19. Hanifin JM, Schneider LC, Leung DY, Ellis CN, Jaffe HS, Izu AE, et al. Recom- J, et al. The long-term safety and efficacy of cyclosporin in severe refractory
binant interferon gamma therapy for atopic dermatitis. J Am Acad Dermatol atopic dermatitis: a comparison of two dosage regimens. Br J Dermatol 1996;
1993;28:189-97. 135(Suppl. 48):15-20.
20. Harper JI, Ahmed I, Barclay G, Lacour M, Hoeger P, Cork MJ, et al. Cyclosporin 44. Zurbriggen B, Wuthrich B, Cachelin AB, Wili PB, Kagi MK. Comparison of two
for severe childhood atopic dermatitis: short course versus continuous therapy. formulations of cyclosporin A in the treatment of severe atopic dermatitis. A
Br J Dermatol 2000;142:52-8. double-blind, single-centre, cross-over pilot study. Dermatology 1999;198:56-60.
21. Heddle RJ, Soothill JF, Bulpitt CJ, Atherton DJ. Combined oral and nasal beclo- 45. GRADE working group. Available at: http://www.gradeworkinggroup.org/
methasone diproprionate in children with atopic eczema: a randomised controlled publications. Accessed October 17, 2013.
trial. Br Med J (Clin Res Ed) 1984;289:651-4. 46. Akdis CA, Akdis M, Bieber T, Bindslev-Jensen C, Boguniewicz M, Eigenmann P,
22. Hon KL, Leung TF, Ng PC, Lam MC, Kam WY, Wong KY, et al. Efficacy and et al. Diagnosis and treatment of atopic dermatitis in children and adults: Euro-
tolerability of a Chinese herbal medicine concoction for treatment of atopic pean Academy of Allergology and Clinical Immunology/American Academy of
438 ROEKEVISCH ET AL J ALLERGY CLIN IMMUNOL
FEBRUARY 2014

Allergy, Asthma and Immunology/PRACTALL Consensus Report. J Allergy Clin 51. Graham-Brown R. Toxicity of Chinese herbal remedies. Lancet 1992;340:
Immunol 2006;118:152-69. 673-4.
47. Hanifin JM, Cooper KD, Ho VC, Kang S, Krafchik BR, Margolis DJ, et al. 52. Hon KL, Chan BC, Leung PC. Chinese herbal medicine research in eczema treat-
Guidelines of care for atopic dermatitis, developed in accordance with the Amer- ment. Chin Med 2011;6:17.
ican Academy of Dermatology (AAD)/American Academy of Dermatology As- 53. MacGregor FB, Abernethy VE, Dahabra S, Cobden I, Hayes PC. Hepatotoxicity
sociation Administrative Regulations for Evidence-Based Clinical Practice of herbal remedies. BMJ 1989;299:1156-7.
Guidelines. J Am Acad Dermatol 2004;50:391-404. 54. Mostefa-Kara N, Pauwels A, Pines E, Biour M, Levy VG. Fatal hepatitis after
48. Leung TF, Wong KY, Wong CK, Fung KP, Lam CW, Fok TF, et al. In vitro and herbal tea. Lancet 1992;340:674.
clinical immunomodulatory effects of a novel Pentaherbs concoction for atopic 55. Ring J, Alomar A, Bieber T, Deleuran M, Fink-Wagner A, Gelmetti C, et al.
dermatitis. Br J Dermatol 2008;158:1216-23. Guidelines for treatment of atopic eczema (atopic dermatitis) part I. J Eur
49. Ring J, Alomar A, Bieber T, Deleuran M, Fink-Wagner A, Gelmetti C, et al. Acad Dermatol Venereol 2012;26:1045-60.
Guidelines for treatment of atopic eczema (atopic dermatitis) Part II. J Eur 56. Saeki H, Furue M, Furukawa F, Hide M, Ohtsuki M, Katayama I, et al. Guidelines
Acad Dermatol Venereol 2012;26:1176-93. for management of atopic dermatitis. J Dermatol 2009;36:563-77.
50. Schneider L, Tilles S, Lio P, Boguniewicz M, Beck L, LeBovidge J, et al. Atopic 57. Williams HC. Clinical practice. Atopic dermatitis. N Engl J Med 2005;352:
dermatitis: a practice parameter update 2012. JAllergy Clin Immunol 2013;131:295-9. 2314-24.