REVIEW

Systematic review and meta-analysis of randomized

clinical trials (RCTs) comparing topical calcineurin
inhibitors with topical corticosteroids for atopic
dermatitis: A 15-year experience
Joris A. Broeders, MD, PhD,a Usama Ahmed Ali, MD,b and Gayle Fischer, MDa
Sydney, Australia, and Utrecht, The Netherlands

Background: Calcineurin inhibitors are alternatives to corticosteroid for treatment of atopic dermatitis.

Objectives: We sought to compare the beneficial effects and adverse events associated with these therapies
in treating patients with atopic dermatitis.

Methods: Four databases were searched for randomized clinical trials comparing topical calcineurin
inhibitors versus corticosteroids in children and adults. Methodological quality was evaluated to assess bias
risk. Clinical outcome and costs were compared.

Results: Twelve independent randomized clinical trials comparing calcineurin inhibitors (n = 3492) versus
corticosteroids (n = 3462) were identified. Calcineurin inhibitors and corticosteroids had similar rates of
improvement of dermatitis (81% vs 71%; risk ratio [RR] 1.18; 95% confidence interval [CI] 1.04-1.34; P = .01)
and treatment success (72% vs 68%; RR 1.15; 95% CI 1.00-1.31; P = .04). Calcineurin inhibitors were
associated with higher costs and had more adverse events (74% vs 64%; RR 1.28; 95% CI 1.05-1.58; P = .02)
including a higher rate of skin burning (30% vs 9%; RR 3.27; 95% CI 2.48-4.31; P \ .00001) and pruritus
(12% vs 8%; RR 1.49; 95% CI 1.24-1.79; P \.00001). There were no differences in atrophy, skin infections,
or adverse events that were serious or required discontinuation of therapy.

Limitations: Only a small number of trials reported costs.

Conclusion: Calcineurin inhibitors and corticosteroids have similar efficacy. Calcineurin inhibitors are
associated with higher costs and have more adverse events, such as skin burning and pruritus. These results
provide level-1a support for the use of corticosteroids as the therapy of choice for atopic dermatitis. ( J Am
Acad Dermatol http://dx.doi.org/10.1016/j.jaad.2016.02.1228.)

Key words: atopic dermatitis; calcineurin inhibitors; corticosteroids; meta-analysis; randomized controlled
trial; systematic review.

A topic dermatitis is a common inflammatory

skin disorder characterized by itch, xerosis,
and inflammation. Superinfection with
Staphylococcus aureus resulting in edema and
Abbreviations used:
CI:
FDA:
RCT:
confidence interval
Food and Drug Administration
randomized clinical trial
exudation is common. These symptoms interfere RR: risk ratio
with most aspects of daily life. Topical corticosteroids
remain the mainstay of treatment of atopic dermatitis.

From the Department of Dermatology, Royal North Shore Reprint requests: Joris A. Broeders, MD, PhD, Royal North Shore
Hospital, Sydneya; and Department of Surgery, University Hospital, Pacific Hyy, St Leonards, NSW 2065, Australia. E-mail:
Medical Center Utrecht.b [email protected].
Funding sources: None. Published online May 11, 2016.
Conflicts of interest: None declared. 0190-9622/$36.00
Accepted for publication February 23, 2016. Ó 2016 by the American Academy of Dermatology, Inc.
http://dx.doi.org/10.1016/j.jaad.2016.02.1228

1
2 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

In 2000, topical immunomodulation by calcineurin perception that they have lower efficacy than corti-
inhibition was introduced to provide an alternative to costeroids.6 Consequently, it is currently unknown
1
topical corticosteroids. Inhibition of calcineurin whether calcineurin inhibitors or corticosteroids
decreases the activity of transcription factors that should be considered as standard therapy for atopic
control cell division, which reduces inflammation by dermatitis. Previous meta-analyses have pooled
means of selective prevention of T-cell activation. studies published up until 2009 and since then 7
Pimecrolimus and tacrolimus are the calcineurin individual randomized clinical trials (RCTs) have
inhibitors that are currently been published. The current
being used in the treatment study aims to bring the
of atopic dermatitis in chil- CAPSULE SUMMARY evidence base up to date
dren and adults. These drugs and determine the therapy
d Calcineurin inhibitors were introduced as
enjoyed an initial surge of of choice for atopic derma-
an alternative to corticosteroid for
popularity fueled by aggres- titis by comparing clinical
treatment of atopic dermatitis.
sive advertising centered on outcome and costs of topical
the claim that, unlike topical d Calcineurin inhibitors and corticosteroids calcineurin inhibitors with
corticosteroids, they would have similar efficacy. Calcineurin corticosteroids.
not produce cutaneous atro- inhibitors are associated with higher
phy. This was dampened by costs and have more adverse events. METHODS
the release of a Food and d The findings provide level-1a support for Study selection
Drug Administration (FDA) use of corticosteroids as the therapy of A systematic literature
Black Box warning. In 2005, choice for atopic dermatitis. search with predefined
the FDA raised concerns search terms (Fig 1) was
about the widespread use of carried out in MEDLINE
525,000 scripts (14% of prescriptions) for pimecroli- (from 1960), EMBASE (from 1980), Cochrane
mus and 69,000 scripts (7% of prescriptions) for Library (issue 1, 2012), and the Web of Knowledge
tacrolimus,2 particularly in regard to off-label pre- Conference Proceedings Citation IndexeScience
scribing to children and heavy direct-to-consumer (from 1990) databases on April 5, 2015 (Fig 1). All
advertising. Based on this, high-dose animal studies, identified articles were screened for cross-
and malignancy reports in the FDA’s adverse event references. Language restrictions were not applied.
reporting system (25 malignancies, 13 lymphomas in
[6.7 million potential patients), the committee Inclusion criteria
concluded that calcineurin inhibitors carry a possible Title and abstract of all identified articles were
and plausible risk of the development of melanoma screened and selected according to the following
skin cancer, nonmelanoma skin cancer, and Hodgkin inclusion criteria: study populationepatients with
and non-Hodgkin lymphoma.3 atopic dermatitis; interventioneclearly documented
In response to the warning, many dermatology topical treatment with calcineurin inhibitors or
associations have released position statements corticosteroids; study outcomeseat least 1 of
promoting the safety of calcineurin inhibitors and, the outcome measures reported below; study
since then, no evidence has been published designepatients assigned to either calcineurin
demonstrating a causal relationship with the inhibitors or corticosteroids by random allocation;
development of malignancy in human beings. and publicationepublished as a full article in a
Calcineurin inhibitors have been used in clinical peer-reviewed journal.
practice for 15 years, but there is ongoing uncertainty
about their role in the treatment of atopic dermatitis. Exclusion criteria
Several randomized trials have compared clinical Studies were excluded from analysis if they did
outcome of calcineurin inhibitors and corticosteroids not meet the inclusion criteria and it was impossible
for atopic dermatitis. Meta-analyses of these trials to extract or calculate appropriate data from the
have so far only evaluated clinical outcomes and published results. Abstracts of RCTs were excluded
concluded that calcineurin inhibitors are as effective as details on treatment, methodological quality, and
as corticosteroids.4-6 Calcineurin inhibitors have a the risk of bias of these studies could not be assessed.
substantial cost premium,7 but meta-analyses have
so far failed to demonstrate superiority over Outcomes of interest and definitions
corticosteroids to justify routine use.4-6 Calcineurin Efficacy outcomes were improvement of
inhibitors are currently not widely used as first-line dermatitis and treatment success. Safety outcome
therapy for atopic dermatitis, because of the included: adverse events, skin burning, pruritus,
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 3

Medline via Pubmed:

"calcineurin inhibitors"[Pharmacological Action] OR "calcineurin inhibitors"[MeSH Terms] OR
"calcineurin inhibitors"[tiab] OR "tacrolimus"[MeSH Terms] OR tacrolimus [tiab] OR Protopic [tiab]
OR Prograf [tiab] OR Advagraf [tiab] OR FK506 [tiab] OR “FK-506” [tiab] OR “FK 506” [tiab] OR
fujimycin [tiab] OR Pimecrolimus [tiab] OR Elidel [tiab] OR “ascomycinmacrolactam” [tiab] OR "SDZ n = 275
ASM 981"[tiab] AND ("dermatitis, atopic"[MeSH Terms] OR ("dermatitis" [tiab] AND "atopic"[ tiab])
OR "atopic dermatitis"[tiab]) AND ((randomized controlled trial [pt] OR controlled clinical trial [pt] OR
randomized [tiab] OR randomised [tiab] OR placebo [tiab] OR clinical trials as topic [mesh: noexp]
OR randomly [tiab] OR trial [ti]) NOT (animals [mh] NOT humans [mh]))

Embase:
'calcineurin'/exp OR (calcineurin AND inhibitors) OR 'tacrolimus'/exp OR tacrolimus OR Protopic
OR Prograf OR Advagraf OR FK506 OR 'FK-506' OR 'FK 506' OR fujimycin OR 'pimecrolimus'/exp
OR pimecrolimus OR Elidel OR 'ascomycinmacrolactam' OR 'SDZ ASM 981'AND (atopic AND
n = 441
('dermatitis'/exp OR dermatitis)) AND (random*:ab,ti OR placebo*:de,ab,ti OR (double NEXT/1
blind*):ab,ti)

Cochrane central:
"calcineurin inhibitors" OR tacrolimus OR Protopic OR Prograf OR Advagraf OR FK506 OR “FK-
506” OR “FK 506” OR fujimycin OR Pimecrolimus OR Elidel OR “ascomycinmacrolactam” OR "SDZ n = 241
ASM 981"AND ("dermatitis" AND "atopic") OR "atopic dermatitis"

Web of Science (CPCI-S):

TOPIC: ("calcineurin inhibitors" OR tacrolimus OR Protopic OR Prograf OR Advagraf OR FK506 OR
“FK-506” OR “FK 506” OR fujimycin OR Pimecrolimus OR Elidel OR “ascomycinmacrolactam” OR
"SDZ ASM 981" ) AND (("dermatitis" AND "atopic") OR "atopic dermatitis") OR TITLE: ("calcineurin n = 156
inhibitors" OR tacrolimus OR Protopic OR Prograf OR Advagraf OR FK506 OR “FK-506” OR “FK
506” OR fujimycin OR Pimecrolimus OR Elidel OR “ascomycinmacrolactam” OR "SDZ ASM 981" )
AND (("dermatitis" AND "atopic") OR "atopic dermatitis")

Total (n = 1113)

Papers excluded n = 305

- Automatic filtering double papers n = 305
Papers retrieved after search
(n = 808)
Papers excluded n = 541
- Non RCT n = 541
Publications on RCTs
(n = 267) Papers excluded n = 228
- Manual filtering double papers n = 27
- Reviews n = 42
- Comparison vs. vehicle n = 97
Relevant publications on RCTs - No comparison of calci vs. cortico n = 62
(n = 39)
Papers excluded n = 13
- Abstract of unpublished RCT n = 12
- Quality of life study of included RCT n=1
Potentially appropriate RCTs to be
included in the meta-analysis
(n = 26)
RCTs withdrawn by outcome n = 12
- No outcomes of interest n = 12

RCTs with usable information

by outcome
(n = 14)
Fig 1. Flow chart illustrating details of the search strategy and study selection process according
to the Quality of Reporting of Meta-analyses statement.8-10 CPCI-S, Conference Proceedings
Citation IndexeScience; MeSH, Medical Subject Heading; RCT, randomized clinical trial.
4 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

adverse events related to treatment, adverse events was used for data management and statistical
requiring treatment discontinuation, severe adverse analyses.
events, atrophy, and skin infection. Costs were
evaluated as well. Subgroup analysis for tacrolimus RESULTS
and pimecrolimus was performed if 2 or more trials Description of studies
could be pooled in meta-analyses. A total of 1113 potential relevant publications
were identified (Fig 1). In all, 267 articles on RCTS
Data extraction were identified, of which 39 compared calcineurin
Titles and abstracts of all retrieved records, and inhibitors versus corticosteroids. Twelve potentially
subsequently full-text articles, were examined relevant RCTs that had been published as an abstract
according to the Quality of Reporting of Meta- only, without a peer-reviewed publication, were
analyses guidelines.8-10 The following data were excluded. One article was excluded because it was
extracted separately for all studies meeting the a quality-of-life study on an RCT that was already
inclusion criteria: reference of study, study included. A further 12 RCTs were excluded because
population characteristics, study design, inclusion they did not report clinical outcomes, but merely
and exclusion criteria, and number of participating reported on physiologic outcome. Finally, 14 articles
subjects for each end point. For each outcome on 13 original RCTs were identified.
the study population and number of events was The included trials were published between 2001
recorded. and 2015. A total of 3492 patients were randomized
to calcineurin inhibitors and 3462 patients to
Risk of bias assessment corticosteroids. Mean follow-up was 101 weeks
Risk of bias was assessed of all articles using both (range 2-260 weeks). All participants applied
the Cochrane Collaboration tool for assessing risk of calcineurin inhibitors or corticosteroids twice daily
bias11 and the Jadad scoring system.12 and all but 1 trial were funded by pharmaceutical
companies (Table I; available at http://www.jaad.
Statistical analysis org). Mean age, sex, and percentage affected body
Statistical analyses were performed following the surface area were divided equally between both
recommendations of the Cochrane Collaboration arms. Most studies included patients with moderate
and Quality of Reporting of Meta-analyses guide- to severe atopic dermatitis (Table II; available at
lines.8-10 Outcomes reported by 3 or more studies http://www.jaad.org).
were pooled in meta-analyses. Outcomes were
presented as risk ratios (RRs) and data were pooled Methodological quality of included studies
using the Mantel-Haenszel method. Trials with The trials had good methodological quality, with a
0 events in both arms were excluded from mean Jadad score of 4 (range 2-5) (Table III; available
meta-analysis. Trials with 0 events in 1 arm were at http://www.jaad.org). Six trials reported the
included in the analysis by adding a continuity method of sequence generation and 1 trial reported
correction of 0.5 to all cells in the 2 3 2 table of allocation concealment. Eleven trials were double-
that study. As a robustness assessment, meta- blind and all trials had adequate report on loss to
analyses with RCTs with 0 events in 1 arm were follow-up and were free of other sources of bias.
also performed using risk differences in a sensitivity
analysis. For all analyses the 95% confidence interval Efficacy
(CI) was calculated. Calcineurin inhibitors and corticosteroids had a
Heterogeneity was calculated using the x2 test of similar percentage of patients with improvement of
Higgins and Thompson,13 and inconsistency in dermatitis (81% vs 71%; RR 1.18; 95% CI 1.04-1.34;
study effects was quantified by I2 values.11,14 The P = .01) (Fig 2) and treatment success (72% vs 68%;
fixed-effects model was used if no heterogeneity was RR 1.15; 95% CI 1.00-1.31; P = .04) (Fig 3). Subgroup
present (x2 P value [ .1 and I2 \ 50%). If excessive analysis with stratification of tacrolimus and
heterogeneity was present, data were first pimecrolimus were in line with these results. An
re-checked and the DerSimonian and Laird15 additional subgroup analysis after exclusion of a
random-effects model was used when heterogeneity study that evaluated a mid-class steroid yielded
persisted. Funnel plots were used to help similar results.
identify the presence of publication or other types
of bias.16-18 Review Manager software (RevMan v. Safety
5.0.16, provided by the Cochrane Collaboration, the The number of adverse events (74% vs 64%; RR
Nordic Cochrane Centre, Copenhagen, Denmark) 1.28; 95% CI 1.05-1.58; P = .02) (Fig 4) and adverse
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 5

Fig 2. Improvement of dermatitis. Please see Table I for reference citations. CI, Confidence
interval; M-H, Mantel-Haenszel.

Fig 3. Treatment success. Please see Table I for reference citations. CI, Confidence interval;
M-H, Mantel-Haenszel.

Fig 4. Adverse events. Please see Table I for reference citations. CI, Confidence interval; M-H,
Mantel-Haenszel.
6 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

Fig 5. Adverse events related to treatment. Please see Table I for reference citations.
CI, Confidence interval; M-H, Mantel-Haenszel.

Fig 6. Skin burning. Please see Table I for reference citations. CI, Confidence interval; M-H,
Mantel-Haenszel.

events related to treatment (11% vs 8%; RR 1.45; 95% Cost

CI 1.15-1.83; P = .002) (Fig 5) were higher in the Costs were only reported by 2 trials and could not
calcineurin inhibitor group compared with the be pooled in meta-analysis. Calcineurin inhibitors
corticosteroid group, with a higher rate of skin were associated with higher medication costs in
burning (30% vs 9%; RR 3.27; 95% CI 2.48-4.31; the first RCT (V101 vs V15). The second RCT
P \.00001) (Fig 6) and pruritus (12% vs 8%; RR 1.49; reported higher total medical costs compared with
95% CI 1.24-1.79; P \ .00001) (Fig 7). There were corticosteroids as well, both for patients with
no differences in adverse events requiring moderate (₤527 vs ₤177) and severe (₤621 vs ₤215)
discontinuation (1.8% vs 1.9%; RR 0.95; 95% CI atopic dermatitis.
0.66-1.38; P = .79) (Fig 8), severe adverse events
(8.2% vs 7.2%; RR 1.15; 95% CI 0.98-1.34; P = .08) DISCUSSION
(Fig 9), atrophy (0.8% vs 0%; RR 5.66; 95% CI Topical corticosteroids are the mainstay of
1.00-31.91; P = .05) (Fig 10), or skin infection (12% treatment of atopic dermatitis. The calcineurin
vs 11%; RR 1.08; 95% CI 0.94-1.24; P = .29) (Fig 11). inhibitors pimecrolimus and tacrolimus were
Funnel plots did not demonstrate evidence of developed as alternatives. Several randomized trials
publication bias, with the largest studies plotted have compared clinical outcome of calcineurin
near the average and smaller studies spread evenly inhibitors and corticosteroids for atopic dermatitis.
on both sides of the average (Fig 12). Subgroup Meta-analyses of these trials have concluded that
analysis after stratification of tacrolimus and calcineurin inhibitors are as effective as corticoste-
pimecrolimus were in line with the main analysis, roids, but have so far failed to demonstrate
but the difference in adverse events did not reach superiority over corticosteroids to justify routine
significance in the pimecrolimus group. An use.4-6 Previous reviews have pooled studies
additional subgroup analysis after exclusion of a published up until 2009 and since then, 7 individual
study that evaluated a mid-class steroid yielded RCTs have been published. The current study aims to
similar results. generate a high level of evidence to determine
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 7

Fig 7. Pruritus. Please see Table I for reference citations. CI, Confidence interval; M-H,
Mantel-Haenszel.

Fig 8. Adverse events requiring treatment discontinuation. Please see Table I for reference
citations. CI, Confidence interval; M-H, Mantel-Haenszel.

Fig 9. Severe adverse events. Please see Table I for reference citations. CI, Confidence interval;
M-H, Mantel-Haenszel.
8 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

Fig 10. Atrophy. Please see Table I for reference citations. CI, Confidence interval; M-H,
Mantel-Haenszel.

Fig 11. Skin infection. Please see Table I for reference citations. CI, Confidence interval; M-H,
Mantel-Haenszel.

with a mean Jadad score of 4. Eleven out of 13 trials

were double-blind and all trials apart from 3
physiologic studies had adequate report on loss to
follow-up. The method of sequence generation and
allocation concealment were rarely reported. All
trials were free of other sources of bias and reported
loss to follow-up adequately.
A benchmark meta-analysis based on 7 RCTs
published up to 2004 concluded that there is little
evidence on how calcineurin inhibitors compare
with corticosteroids and recommended RCTs
with longer follow-up.5 Since then, 2 RCTs with a
follow-up of 2 months and 3 RCTs with a follow-up
Fig 12. Funnel plot of severe adverse events. RR, Risk
ratio.
of more than 6 months have been published and
were included in the current meta-analysis. The
present meta-analysis includes 13 RCTs with a
choice of therapy for atopic dermatitis by comparing mean duration of follow-up of 2 years. Individual
both clinical outcome and costs of topical calcineurin RCTs have concluded that outcome after calcineurin
inhibitors to corticosteroids. inhibitor and corticosteroid treatment is similar.
The current study is the largest meta-analysis on This could potentially be because of a type 2
this topic with 6954 children and adults with error and therefore it is valuable to increase the
moderate to severe atopic dermatitis included. All power of statistical analysis by pooling the results of
participants applied topical therapy twice daily. these RCTs in meta-analyses. The power of the
Patient populations were comparable among trials current meta-analysis is substantial with a patient
and divided equally between the arms of each trial. population of 6954 subjects and demonstrates that
The methodological quality of the 13 RCTs was good, there was a statistical difference in treatment success
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 9

and improvement of dermatitis in favor of corticosteroids remain the treatment of choice for
calcineurin inhibitors, but this was not clinically atopic dermatitis and strengthens this conclusion.
significant compared with corticosteroids. It should These results provide level-1a support for the
however be noted that 8 of the available 12 RCTs continued use of corticosteroids as the therapy of
compared calcineurin inhibitors with low-potency choice for atopic dermatitis. Calcineurin inhibitors
topical corticosteroids, which introduced a bias remain suitable as adjunctive therapy to be used with
toward higher efficacy in the calcineurin inhibitor topical steroids for atopic dermatitis.
group.
The number of adverse events and adverse events REFERENCES
related to treatment was higher in the calcineurin 1. Reitamo S, Wollenberg A, Sch€ opf E, et al. Safety and efficacy of
inhibitor group compared with the corticosteroid 1 year of tacrolimus ointment monotherapy in adults with
group, with a higher rate of skin burning and atopic dermatitis: The European Tacrolimus Ointment Study
Group. Arch Dermatol. 2000;136:999-1006.
pruritus. These adverse events were not severe 2. Briefing document from Patty Greene, Drug Use Data
enough to lead to a difference in treatment Analyst, Division of Epidemiology, Office of Surveillance and
discontinuation. There were no differences in severe Epidemiology, FDA Center for Drug Evaluation and Research.
adverse events or atrophy. Infections were of BPCA drug use review: comparison of ElidelÒ cream and
ProtopicÒ ointment utilization trends following 2006 labeling
particular interest because immunomodulators can
changes. Available from: URL: http://www.fda.gov/downloads/
potentially lead to local immunosuppression, but no AdvisoryCommittees/CommitteesMeetingMaterials/Pediatric
difference in skin infections were identified. AdvisoryCommittee/UCM204723.pdf. 2009. Accessed April 5,
A comparison of calcineurin inhibitors with 2015.
vehicle was not included in the current review 3. Chesney J. Summary Minutes of the Pediatric Advisory
because such an evaluation does not reflect decision Committee. Available from: URL: http://www.fda.gov/ohrms/
dockets/ac/05/minutes/2005-4089m1_Minutes.pdf. 2005. Ac-
making. The internal validity of the current study is cessed April 5, 2015.
high because the analysis was based on high-quality 4. Svensson A, Chambers C, Ganemo A, Mitchell SA. A systematic
RCTs, of adequate size, with high follow-up rates and review of tacrolimus ointment compared with corticosteroids
low risk of bias. The trials were performed across 5 in the treatment of atopic dermatitis. Curr Med Res Opin. 2011;
continents and in both children and adults, which 27:1395-1406.
5. Ashcroft DM, Dimmock P, Garside R, Stein K, Williams HC.
increases the external validity of this meta-analysis. Efficacy and tolerability of topical pimecrolimus and
A limitation of the current meta-analysis is the small tacrolimus in the treatment of atopic dermatitis:
number of trials that reported costs. However, meta-analysis of randomized controlled trials. BMJ. 2005;330:
medication and total medical costs associated with 516.
6. El Batawy MM, Bosseila MA, Mashaly HM, Hafez VS. Topical
calcineurin inhibitors were about 3 times higher than
calcineurin inhibitors in atopic dermatitis: a systematic review
corticosteroids, indicating a large effect size between and meta-analysis. J Dermatol Sci. 2009;54:76-87.
both treatment groups. A further limitation is that the 7. Garside R, Stein K, Castelnuovo E, et al. The effectiveness and
current results were obtained within strict protocols cost-effectiveness of pimecrolimus and tacrolimus for atopic
and follow-up of RCTs. In real-world clinical eczema: a systematic review and economic evaluation. Health
practice, particularly in the long-term, adherence to Technol Assess. 2005;29:1-230.
8. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF.
therapy may well be poorer. The current review Improving the quality of reports of meta-analyses of
pools studies that treated atopic dermatitis of the randomized controlled trials: the QUOROM statement.
entire body; it would be valuable to perform an RCT Quality of Reporting of Meta-analyses. Lancet. 1999;354:
with focus on application of calcineurin inhibitors 1896-1900.
and corticosteroids to the face only, with specific 9. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of
observational studies in epidemiology: a proposal for
outcomes such as glaucoma and cataracts. reporting. Meta-analysis Of Observational Studies in
In conclusion, calcineurin inhibitors and cortico- Epidemiology (MOOSE) group. JAMA. 2000;283:2008-2012.
steroids have similar efficacy for atopic dermatitis. 10. Clarke M, Horton R. Bringing it all together: Lancet-Cochrane
Calcineurin inhibitors are associated with higher collaborate on systematic reviews. Lancet. 2001;357:1728.
costs and have more adverse events, such as skin 11. Higgins JP, Green S. Cochrane handbook for systematic reviews
of interventions. Chichester: John Wiley and Sons; 2008.
burning and pruritus. Contrary to popular opinion, 12. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of
there is no evidence that they represent a better reports of randomized clinical trials: is blinding necessary?
alternative regarding the possible adverse effect of Control Clin Trials. 1996;17:1-12.
cutaneous atrophy. The bias toward calcineurin 13. Higgins JP, Thompson SG. Quantifying heterogeneity in a
inhibitors introduced by the fact that the majority of meta-analysis. Stat Med. 2002;21:1539-1558.
14. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring
trials compared tacrolimus 0.1% and 0.03% with inconsistency in meta-analyses. BMJ. 2003;327:557-560.
corticosteroids of inherently less anti-inflammatory 15. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control
potency did not influence the conclusion that Clin Trials. 1986;7:177-188.
10 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

16. Begg CB, Mazumdar M. Operating characteristics of a rank in adults with moderate to severe atopic dermatitis: a random-
correlation test for publication bias. Biometrics. 1994;50:1088-1101. ized, double-blind, comparative trial. Acta Derm Venereol. 2010;
17. Egger M, Davey SG, Schneider M, Minder C. Bias in 90:170-174.
meta-analysis detected by a simple, graphical test. BMJ. 26. Reitamo S, Rustin M, Ruzicka T, et al. Efficacy and safety of
1997;315:629-634. tacrolimus ointment compared with that of hydrocortisone
18. Macaskill P, Walter SD, Irwig L. A comparison of methods to detect butyrate ointment in adult patients with atopic dermatitis.
publication bias in meta-analysis. Stat Med. 2001;20:641-654. J Allergy Clin Immunol. 2002;109:547-555.
19. Bieber T, Vick K, Folster-Holst R, et al. Efficacy and safety of 27. Reitamo S, Van Leent EJM, Ho V, et al. Efficacy and safety of
methylprednisolone aceponate ointment 0.1% compared to tacrolimus ointment compared with that of hydrocortisone
tacrolimus 0.03% in children and adolescents with an acute acetate ointment in children with atopic dermatitis. J Allergy
flare of severe atopic dermatitis. Allergy Eur J Allergy Clin Clin Immunol. 2002;109:539-546.
Immunol. 2007;62:184-189. 28. Reitamo S, Harper J, Dbos J, et al. 0.03% Tacrolimus ointment
20. Doss N, Reitamo S, Dubertret L, et al. Superiority of tacrolimus applied once or twice daily is more efficacious than 1%
0.1% ointment compared with fluticasone 0.005% in adults hydrocortisone acetate in children with moderate to severe
with moderate to severe atopic dermatitis of the face: results atopic dermatitis: results of a randomized double-blind
from a randomized, double-blind trial. The British journal of controlled trial. Br J Dermatol. 2004;150:554-562.
dermatology. 2009;161:427-434. 29. Reitamo S, Ortonne JP, Sand C, et al. A multicenter,
21. Doss N, Kamoun M-R, Dubertret L, et al. Efficacy of tacrolimus randomized, double-blind, controlled study of long-term
0.03% ointment as second-line treatment for children treatment with 0.1% tacrolimus ointment in adults with
with moderate-to-severe atopic dermatitis: evidence from moderate to severe atopic dermatitis. Br J Dermatol. 2005;
a randomized, double-blind non-inferiority trial vs. flutica- 152:1282-1289.
sone 0.005% ointment. Pediatr Allergy Immunol. 2010;21:321-329. 30. Hjelmgren J, Svensson A, Jorgensen ET, Lindemalm-
22. Hofman T, Cranswick N, Kuna P, et al. Tacrolimus ointment Lundstam B, Ragnarson TG. Cost-effectiveness of tacrolimus
does not affect the immediate response to vaccination, the ointment vs. standard treatment in patients with moderate
generation of immune memory, or humoral and cell-mediated and severe atopic dermatitis: a health-economic model
immunity in children. Arch Dis Child. 2006;91:905-910. simulation based on a patient survey and clinical trial data.
23. Luger T, Van Leent EJM, Graeber M, et al. SDZ ASM 981: an Br J Dermatol. 2007;156:913-921.
emerging safe and effective treatment for atopic dermatitis. Br 31. Sigurgeirsson B, Boznanski A, Todd G, et al. Safety and efficacy
J Dermatol. 2001;144:788-794. of pimecrolimus in atopic dermatitis: a 5-year randomized trial.
24. Luger TA, Lahfa M, Folster-Holst R, et al. Long-term safety and Pediatrics. 2015;135:597-606.
tolerability of pimecrolimus cream 1% and topical corticoste- 32. Sikder M, Al MS, Khan RM, Chowdhury AH, Khan HM,
roids in adults with moderate to severe atopic dermatitis. Hoque MM. Topical 0.03% tacrolimus ointment, 0.05%
J Dermatol Treat. 2004;15:169-178. clobetasone butyrate cream alone and their combination in
25. Mandelin J, Remitz A, Virtanen H, Reitamo S. One-year treatment older children with atopic dermatitis - an open randomized
with 0.1% tacrolimus ointment versus a corticosteroid regimen comparative study. J Pak Assoc Dermatol. 2005;15:304-312.
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 10.e1

Table I. Details of included randomized clinical trials comparing calcineurin inhibitors versus corticosteroids
Author Year Therapy Frequency n Duration, wk Class* Funding
Bieber et al19 2007 Tacrolimus 0.03% TD 136 3 Intendis GmbH
Methylprednisolone 0.1% 129 7
Doss et al20 2009 Tacrolimus 0.1% TD 288 3 Astellas Pharma Ltd
Fluticasone 0.005% 280 5
Doss et al21 2010 Tacrolimus 0.03% TD 240 3 Astellas Pharma Ltd
Fluticasone 0.005% 239 5
Hofman et al22 2006 Tacrolimus 0.03% TD 121 2 Astellas Pharma Ltd
Hydrocortisone acetate 0.1% 111 5
Luger et al23 2001 Pimecrolimus 1% TD 45 3 NR
Betamethasone valerate 0.1% 42 6
Luger et al24 2004 Pimecrolimus 1% TD 328 52 Novartis Pharma AG
Triamcinolone 0.1% 1 330 5
hydrocortisone acetate 7
1% (face)
Mandelin et al25 2010 Tacrolimus 1% TD 40 52 Astellas Pharma Ltd
Hydrocortisone butyrate 40 5
0.1% 1 hydrocortisone 7
acetate 1% (face)
Reitamo et al26 2002 Tacrolimus 0.1% TD 191 3 Fujisawa GmbH
Hydrocortisone butyrate 0.1% 186 5
Reitamo et al27 2002 Tacrolimus 0.1% TD 186 3 Fujisawa GmbH
Hydrocortisone acetate 1% 185 7
Reitamo et al28 2004 Tacrolimus 0.03% TD 210 3 Fujisawa GmbH
Hydrocortisone acetate 1% 207 7
Reitamo et al29-31 2005 Tacrolimus 0.1% TD 487 26 Fujisawa GmbH
Hydrocortisone butyrate 485 5
0.1% 1 hydrocortisone 7
acetate 1% (face)
Sigurgeirsson 2015 Pimecrolimus 1% TD 1205 260 Novartis Pharma
et al31 Hydrocortisone butyrate 1213 5
0.1% or hydrocortisone 7
acetate 1%
Sikder et al32 2005 Tacrolimus 0.03% TD 15 4 Square Pharmaceuticals
Clobetasone 0.05% 15 4

Intendis GmbH, Berlin, Germany; Astellas Pharma Ltd, Munich, Germany; Novartis Pharma AG, Basel, Switzerland; Fujisawa GmbH, Munich,
Germany; Square Pharmaceuticals, Dhaka, Bangladesh.
NR, Not reported; TD, twice daily.
*Potency: 1 = Super; 2 = high; 3 = upper mid; 4 = mid; 5 = lower mid; 6 = low; 7 = least.
10.e2 Broeders, Ahmed Ali, and Fischer J AM ACAD DERMATOL

Table II. Patient characteristics

Author Therapy Range age, y Mean age, y Male/female Severity AD BSA, %
Bieber et al19 Calcineurin 2-15 7.5 NR Severe-very severe 29
Corticosteroid 7.8 NR 29
Doss et al20 Calcineurin Adult 35 143/140 Moderate-severe NR
Corticosteroid 35 117/162
Doss et al21 Calcineurin 2-15 6.9 114/122 Moderate-severe 25
Corticosteroid 6.6 112/125 24
Hofman et al22 Calcineurin 2-11 6.2 58/63 Moderate-severe 27
Corticosteroid 6.0 47/64 32
Luger et al23 Calcineurin Adult 28 24/21 $Moderate NR
Corticosteroid 32 19/23
Luger et al24 Calcineurin Adult 33 146/182 Moderate-severe 27
Corticosteroid 34 153/177 27
Mandelin et al25 Calcineurin Adult 29 9/31 Moderate-severe 58
Corticosteroid 29 10/30 62
Reitamo et al26 Calcineurin Adult 32 82/109 Moderate-severe 30
Corticosteroid 31 87/99 36
Reitamo et al27 Calcineurin 2-15 7.2 96/90 Moderate-severe 23
Corticosteroid 7.2 95/90 25
Reitamo et al28 Calcineurin 2-15 6.9 95/115 Moderate-severe 37
Corticosteroid 7.2 107/100 39
Reitamo et al29-31 Calcineurin Adult 32 225/262 Moderate-severe 36
Corticosteroid 33 224/261 38
Sigurgeirsson et al31 Calcineurin .25-1 0.6 744/461 Mild-moderate 21
Corticosteroid 0.6 742/471 21
Sikder et al32 Calcineurin 7-15 10 6/9 Moderate-severe 25
Corticosteroid 11 11/4 25

AD, Atopic dermatitis, BSA, body surface area; NR, not reported.
J AM ACAD DERMATOL Broeders, Ahmed Ali, and Fischer 10.e3

Table III. Risk of bias summary

Sequence Allocation Blinding Blinding Report loss No other Jadad
generation concealment observer patient follow-up bias score
Bieber et al19 No No Yes Yes Yes Yes 4
Doss et al20 Yes No Yes Yes Yes Yes 5
Doss et al21 Yes No Yes Yes Yes Yes 5
Hofman et al22 No No Yes Yes Yes Yes 4
Luger et al23 No No Yes Yes Yes Yes 3
Luger et al24 No No Yes Yes Yes Yes 3
Mandelin et al25 No No Yes Yes Yes Yes 3
Reitamo et al26 Yes No Yes Yes Yes Yes 4
Reitamo et al27 Yes No Yes Yes Yes Yes 5
Reitamo et al28 No No Yes Yes Yes Yes 3
Reitamo et al29 Yes Yes Yes Yes Yes Yes 5
Sigurgeirsson et al31 Yes No No No Yes Yes 3
Sikder et al32 No No No No Yes Yes 2