Veterinary Dermatology 2005, 16, 52 60

Treatment of localized lesions of canine atopic dermatitis with

tacrolimus ointment: a blinded randomized controlled trial

Blackwell Publishing, Ltd.

EMMANUEL BENSIGNOR* and THIERRY OLIVRY

*Dermatology Unit, Veterinary Clinic, 75003 Paris and 35510 Cesson-Svign, France
Department of Clinical Sciences, College of Veterinary Medicine, North Carolina State University,
Raleigh, NC 27606, USA
(Received 3 March 2004; accepted 20 June 2004)

Abstract This investigator-blinded randomized controlled trial was designed to determine whether tacrolimus
ointment (Protopic, Fujisawa Healthcare) decreased the severity of localized lesions of canine atopic dermatitis
(AD). Twenty dogs with AD were enrolled if they exhibited lesions on both front metacarpi. Each foot was randomized to be treated with 0.1% tacrolimus or placebo (vaseline) ointment twice daily for 6 weeks. Before, and
every 2 weeks during the study, erythema, lichenification, oozing and excoriations each were graded on a 10-point
scale (maximal total score: 40). The primary outcome measures were the percentage reduction from baseline of
lesional scores and the number of subjects whose scores had decreased by 50% or greater at study end. Intentionto-treat analyses were used. At study onset, lesional scores were not significantly different between sites treated
with tacrolimus or placebo. After 6 weeks, the percentage reduction from baseline scores was higher for tacrolimus-treated sites (median: 63%; 95% confidence interval: 3967) than for placebo-treated feet (median: 3%;
confidence interval: 213) (Wilcoxon test; P = 0.0003). When tacrolimus was applied, lesions decreased by 50%
or greater in 15/20 dogs (75%); these dogs were those that completed the study. In contrast, this benchmark was
not reached for any placebo-treated feet (Fishers test; P < 0.0001). Adverse drug events consisted of minor irritation in some lesional areas treated with tacrolimus. Results of this trial suggest that the application of 0.1% tacrolimus ointment is useful for reducing the severity of localized skin lesions of canine AD.

I N TRO D U CTI ON
Canine atopic dermatitis (AD) is defined as a genetically
predisposed inflammatory and pruritic allergic skin
disease with characteristic clinical features.1 Although
no reliable prevalence and incidence data are available,
AD is considered a skin disease seen commonly by
general practitioners and veterinary dermatologists.2
In dogs, the alleviation of clinical signs of AD
requires a multifaceted approach that includes allergen
avoidance, allergen-specific immunotherapy as well as
antimicrobial and anti-inflammatory pharmacotherapies.3 A recent systematic review evaluated the evidence of efficacy of pharmacological interventions for
canine AD.4 This review of published clinical trials
concluded that there is good evidence for recommending the use of oral glucocorticoids and cyclosporin for
treatment of canine AD, and fair evidence for using
topical triamcinolone spray, topical tacrolimus lotion,
oral pentoxifylline or oral misoprostol.4
Tacrolimus is a macrolide immunomodulator that is
synthesized by the fungus Streptomyces tsukubaensis.
Tacrolimus diffuses across cell membranes and binds
to ubiquitous peptidyl-prolyl cis-trans isomerases
designated as FK506 (tacrolimus)-binding proteins
Correspondence to: T. Olivry, Department of Clinical Sciences,
North Carolina State University, College of Veterinary Medicine,
4700 Hillsborough Street, Raleigh, NC 27606, USA. E-mail:
thierry_olivry@ncsu.edu
52

(FKBP, macrophilin-12). Binding of tacrolimus to

FKBP prevents the ability of the calcineurin phosphatase to dephosphorylate a cytoplasmic protein
(NFAT) that activates the transcription of interleukin2 (IL-2) mRNA, thus the translation and secretion
of IL-2 are markedly reduced.5 Tacrolimus also inhibits the transcription of genes encoding other proinflammatory cytokines such as IL-3, IL-4, IL-5,
interferon gamma, tumour necrosis factor alpha and
granulocyte-macrophage colony-stimulating factor.5
The inhibition of cytokine gene transcription leads to
secondary inactivation of the function of most immune
cells involved in the allergic response, such as lymphocytes, dendritic cells, mast cells and eosinophils.6
A recent systematic review evaluated the efficacy of
tacrolimus ointment for treatment of atopic dermatitis
in humans.7 Seven randomized controlled trials
(RCTs), published between 1997 and 2002, were analysed in detail. These trials had enrolled more than
3000 adults and children with AD. Topical tacrolimus,
at both 0.03% and 0.1% concentrations, was found to
be effective when compared to vehicle in five RCTs.
Topical tacrolimus was deemed superior in efficacy
compared to a very weak topical glucocorticoid (1%
hydrocortisone), and the 0.1% tacrolimus ointment
appeared to be equivalent in potency to that of potent
topical glucocorticoid formulations (hydrocortisone
butyrate or betamethasone valerate). Tacrolimus ointment was found to be safe in the short term, even
though a transient burning sensation was felt by half of
2005 European Society of Veterinary Dermatology

Tacrolimus for localized atopic dermatitis

adult patients who used it. Topical tacrolimus ointment
is at least 10 times more expensive than beta-methasone
valerate, however.
Because of such consistent efficacy, recent practice
guidelines recommended the use of tacrolimus ointment (0.1% for adults; 0.03% for children) for longterm intermittent therapy of patients not adequately
responsive to, or intolerant of, conventional therapy.8
Proposed dosing was twice daily until at least 1 week
after signs and symptoms had cleared.8 Remarkably,
this recommendation did not take into account the fact
that RCTs published thus far had NOT enrolled subjects with AD refractory to conventional therapy with
topical glucocorticoids.7
In veterinary dermatology, there are only two trials
examining the efficacy of tacrolimus for treatment of
canine AD. A small-sized study evaluated the efficacy
and safety of 0.3% tacrolimus lotion in atopic dogs.9 In
this double-blinded cross-over RCT, eight dogs with
well-characterized AD were enrolled. Dogs were randomized to treatment with either compounded 0.3%
tacrolimus (Prograf, Fujisawa Healthcare, Deerfield, IL,
USA) or placebo (vehicle) lotions once daily for 4 weeks.
After a 2-week wash out, treatments were reversed. At
study end, there were no significant differences in
owner-assessed pruritus scores within treatment
groups (week 4 compared to baseline) or between
interventions. After 4 weeks, and compared to baseline, clinician-assessed pruritus scores were significantly lower in tacrolimus-treated dogs, but not in
those receiving placebo. At the end of the study, investigator-graded erythema scores, but not pruritus
scores, were significantly lower in dogs treated with
tacrolimus compared to those receiving the placebo
vehicle. Using data provided by this studys leading
author, it was calculated that median reductions from
baseline scores were 0, 33 and 50% for owner-assessed
pruritus, and investigator-graded erythema or pruritus, respectively. Thirty-eight per cent of dogs treated
with tacrolimus exhibited a greater than 50% improvement from baseline in erythema scores, while 1350%
of subjects reached the same level of improvement for
owner- and investigator-assessed pruritus, respectively.
Adverse drug events following tacrolimus lotion application were not reported in this paper.
The same investigators recently completed a second
small RCT to test the efficacy of the commercially
available 0.1% tacrolimus ointment (Protopic, Fujisawa
Healthcare).10 Fourteen dogs with AD were enrolled in
a trial designed similarly to the one described above.9
Twelve subjects remained available at the end of the study,
and intention-to-treat analyses were not performed. At
trials end, investigator-assessed lesional scores were
significantly lower than before treatment with tacrolimus, but not with placebo. A greater than 50%
reduction in lesional scores was achieved for 7/12 dogs
(58%) treated with tacrolimus, and 3/12 dogs (25%)
treated with placebo. Of note is that 5/7 dogs (71%)
that had achieved such a benchmark after tacrolimus
application were affected with localized AD. Indeed,

53

the average reduction in lesional scores was higher for

dogs with localized AD (60%) than for patients with
generalized disease (24%). In this study, a greater than
50% reduction in owner-evaluated pruritus values was
observed in 5/12 dogs (41%) during tacrolimus treatment and in 2/12 dogs (16%) treated with placebo.
Finally, results of this trial confirmed that tacrolimus
blood levels remained below the level of toxicity when
assessed during treatment. Moreover, adverse effects
were not observed in any of the dogs treated with this
commercially available tacrolimus formulation.
The purpose of the present study was to test whether
0.1% tacrolimus ointment (Protopic, Fujisawa Healthcare,
La Selle St Cloud, France) was effective for treatment
of localized lesions of AD in dogs using a blinded RCT
design. This trial will provide evid-ence that the application of tacrolimus ointment leads to a relevant
decrease in the severity of skin lesions of canine AD.
Whenever possible, the results of this clinical trial are
reported according to the CONSORT guidelines
(http://www.consort-statement.org page last
accessed on 1 March, 2004).

M AT E R IA L A N D M E T H O D S
Study design
This clinical trial was designed as a mono-centre,
solo-investigator, side-by-side parallel, single investigatorblinded randomized placebo-controlled experiment of
6-week duration.

Study subjects
Inclusion criteria Client-owned dogs with AD were
recruited for this study from the clientele of CessonSvign (Brittany) and Paris in France by one of the
authors (EB). Study participants were selected from
primary practice (14 dogs; 70%) or after referral by veterinary practitioners following failure of standard-ofcare therapy (six subjects; 30%).
In all dogs, the diagnosis of nonseasonal AD was made
on the basis of a compatible history and clinical signs
similar to those proposed by Willemse11 and Prlaud.12
As recommended recently by the ACVD Task Force on
Canine AD, intradermal tests or allergen-specific IgE
serology were not required for diagnosis of this disease.13
To be eligible for enrolment, study subjects needed
to have skin lesions localized to dorsal and/or ventral
metacarpi of both feet. All dogs also exhibited lesions
at other body sites that included hind feet (16/20; 80%),
face (14/20; 70%), concave (inner) pinnae (20/20;
100%), abdomen (12/20; 60%), axillae (13/20; 65%) and
perineum (8/20; 40%).
In addition to characteristic clinical signs, all resembling or coexisting pruritic skin diseases, such as flea
allergy dermatitis, sarcoptic acariasis (scabies) or cutaneous adverse food reactions (allergy or intolerance)
had to be ruled out. These entities were eliminated
from consideration following appropriate flea control

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 5260

54

E Bensignor and T Olivry

measures (monthly fipronil spot-on or spray), acaricidal therapy, or 6 weeks home-made or hydrolysed
commercial restrictive diets, respectively. Finally, all
surface, superficial or deep bacterial infections and
surface Malassezia yeast infections had to be treated
successfully before enrolment.
Prior to the trial, oral and injectable glucocorticoids
had to be discontinued for at least 3 and 6 weeks,
respectively. Similarly, antihistamines could not have
been prescribed for 2 weeks and cyclosporin for
1 month before the study. Topical treatment could not
have been applied for at least 3 days before the study.
When patients met inclusion criteria, detailed information of the study was provided to their owner, and a
consent form was signed.
Exclusion criteria During this study, the occurrence of
an active bacterial or yeast skin infection, the use of
anti-inflammatory or antipruritic drugs by the dogs
owner, and the development of unacceptable adverse
drug effects all justified immediate withdrawal of the
subject from this trial.
Sample size estimation It was determined that at least
20 dogs needed to be included for this trial to have a
95% power of detecting differences of 35% or greater in
percentage reductions of lesional scores between active
and placebo interventions (standard deviation 30%;
P = 0.05, two-sided) (InStat, GraphPad Software, San
Diego, CA, USA).

Intervention
After satisfaction of inclusion criteria, left and right
feet were randomized for treatment with either 0.1%
tacrolimus ointment (Protopic, Fujisawa Healthcare)
or placebo (vaseline, Monot, Brussels, Belgium) ointment twice daily for 6 weeks. Follow-up of subjects and
assessment of lesional scores was carried out after 2, 4
and 6 weeks. To minimize licking of the topical medication, dog owners were asked to walk or carefully
watch their pet for approximately 15 min. If other body
lesions were present, they were left untreated so that
the evaluation of efficacy of tacrolimus or placebo was
not confounded.

who graded skin lesions. Although pet owners were not

blinded to intervention allocation, they did not participate in the evaluation of treatment efficacy. Compliance to treatment was confirmed by questioning owners
during the first visit whether they had applied both
medications as prescribed. Details on the nature of
drugs applied to each foot were however not asked.

Scoring and outcome measures

The clinician-assessed scoring of pedal lesions consisted
of a grading of erythema, lichenification, oozing and
excoriations on a 10-point linear visual analogue scale.
All four lesion grades were added to yield a total lesional
score, which had a maximum value of 4 10 = 40.
The primary outcome measure was to compare the
percentage change from baseline of clinician-assessed
lesional scores between tacrolimus and placebo at
study end. The second outcome variable was to determine the proportion of subjects achieving a 50% or
greater reduction in such scores at study end with
either intervention. The halving of lesional values during treatment has been shown in previous AD trials to
be associated with an overall assessment of efficacy by
clinician investigators and dog owners.14,15 Thus, this
benchmark was adopted here as a measure of treatment success.
The rate of treatment success (50% or greater reduction from baseline scores) was determined for both
control (e.g. placebo) and experimentally (e.g. tacrolimus) treated feet. Absolute benefit increase (ABI),
which is defined as the absolute arithmetic difference in
the event rate (e.g. percentage of dogs achieving treatment success) between tacrolimus and placebo-treated
feet), and number needed to treat (NNT), the inverse
of ABI, were calculated according to Cockcroft and
Holmes.16

Evaluation of adverse drug events

The nature, extent, severity and frequency of adverse
events were recorded for active and placebo interventions at all three re-evaluation visits. The reasons for
treatment discontinuation were noted in the subjects
file.

Statistical analyses
Randomization and maintenance of blinding
After the veterinary dermatologist (EB) had verified
that subjects met selection criteria, and after obtaining
owners consent, a veterinary technician randomized
the side of treatment (left or right foot) with tacrolimus
and placebo by flipping a coin. Treatment allocation
was concealed from the evaluating clinician, with the
randomization sequence being kept by the technician.
A prescription for Protopic 0.1% and vaseline ointments was given by the technician to the owner who,
after purchase from a local pharmacy, applied both
drugs to sides determined by the previous coin toss.
The clinician remained unaware of treatment allocation sides until each subject had completed the trial.
Therefore, this trial was masked solely to the investigator

Outcome measures were assessed using an intentionto-treat (ITT) analysis. Lesional values of all subjects
were taken into account for statistical analyses,
whether or not the patient had completed the trial. In
this study, five of 20 subjects were no longer available
at trials end (see below for reasons). All missing
lesional scores were replaced using the last value carry
forward rule.
Statistical analyses were performed using Prism software (v. 4.0 for Macintosh, GraphPad software). Nonparametric tests were employed for all analyses. For
comparison of data within treatment groups, repeatedmeasures was used (Friedman test). For comparison of values between tacrolimus and placebo
interventions, before and during treatment, Wilcoxon

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 52 60

Tacrolimus for localized atopic dermatitis

55

signed rank tests were used. Contingency tables to

compare the number of subjects achieving 50% or
greater reduction in scores were analysed with Fishers
exact tests. Two-tailed assessments were used for all
analyses, and values of P < 0.05 were considered
significant. Whenever possible, data were reported as
median (95% confidence interval [CI]).

At baseline, pedal lesional scores were not significantly different between tacrolimus-(median: 28 units;
95% CI: 2430) and placebo-treated feet (median: 26
units: 95% CI: 2428) (Wilcoxon signed rank test;
P = 0.221) (Table 1). Thus, it was felt that the a priori
randomization of treatment allocation between feet
had been effective.

Conflict of interest and funding of the study

RESU LTS

Accounting of study subjects Fifteen of 20 dogs completed the study (Fig. 1). Three dogs (#5, 9 and 18)
were enrolled in this trial, and owners were prescribed
both active and placebo ointments, but they did not
return for re-evaluation of their pets. Owners of two
other dogs (#11 and 14) perceived that both interventions were ineffective, and they requested withdrawal
from the study after 2 and 4 weeks, respectively. However, scores from all 20 subjects were kept for ITT analyses. Missing values were replaced by the last one
recorded which could be the baseline score being
carried forward.

Subjects

Outcome measures

Baseline demographic and clinical characteristics Twenty

dogs were recruited for this trial between May and
September 2003 (Fig. 1). Twelve females and eight
males were selected. Twelve breeds were represented
(four cocker spaniels and West Highland white terriers,
two German shorthaired pointers and French bulldogs, and one each of bull terrier, whippet, German
shepherd, English setter, boxer, basset hound, poodle
and shar pei). The median age of onset of AD was
19 months (range: 648 months), and the median
duration of disease prior to study enrolment was
10 months (range: 1 month10 years).

During the trial, lesional scores of feet treated with

tacrolimus decreased significantly compared to baseline (Friedman test: P < 0.0001) (Table 1; Fig. 2). The
mean of differences between baseline and final (week 6)
scores was 14 units (95% CI: 1018) (Wilcoxon signed
rank test; P = 0.0003), indicating a strong treatment
effect. In contrast, scores of feet on which placebo
(vaseline) had been applied did not decrease significantly during the study (Friedman test: P = 0.156)
(Table 1; Fig. 2). For the placebo group, the mean of
differences between baseline and final scores was only
2 units (95% CI: 03) (Wilcoxon signed rank test;
P = 0.136), thus showing a lack of treatment efficacy.

All drugs used in this trial were paid for by the dogs
owners. The office visits were free of charge. None of
the authors has had any financial ties with the company manufacturing tacrolimus ointment for lecturing
honorarium, funding of research projects, unrestricted
or restricted gifts, or for consulting activities for pay.
Tacrolimus ointment (Protopic) is not currently
approved in any country for treatment of AD in dogs.

Figure 1. Study organization.

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 5260

56

E Bensignor and T Olivry

Table 1. Evolution of pedal lesional scores

Actual lesional scores

Baseline
Week 2
Week 4
Week 6

Tacrolimus

Placebo

Wilcoxon test
P-value

28 (24 30)
20 (1623)
16 (1219)
11 (917)

26 (24 28)
24 (2226)
23 (2226)
24 (2227)

0.2208
0.0032
0.0002
0.0002

Reduction in lesional scores (%)

Baseline
Week 2
Week 4
Week 6

Tacrolimus

Placebo

Wilcoxon test
P-value

30 (1838)
50 (3256)
63 (3967)

8 (013)
7 (113)
3 (213)

0.0004
0.0003
0.0003

Actual lesional scores and reduction of lesional scores are reported as median (95% CI).

Figure 2. Scatter plot of pedal lesional scores during the clinical

trial. Each point represents a study subject. Closed circles are values
from dogs that completed the trials. Open circles correspond to
missing values carried forward because of withdrawal of study
subjects. All data sets were kept for intention-to-treat analyses.

At the trials end, the percentage reduction from

baseline scores was significantly higher for feet treated
with tacrolimus (median: 63%; 95% CI: 3967%) than
for those treated with placebo (median: 3%; 95% CI:
213%) (Wilcoxon signed rank test: P = 0.0003)
(Table 1; Fig. 3a).
After 6 weeks, treatment success, defined by a 50%
or greater reduction from baseline scores, was achieved
in three quarters of the feet treated with tacrolimus,
while this benchmark was not reached for any feet on
which the placebo was applied (Fishers exact test:
P < 0.0001) (Fig. 3b). Significantly, this benchmark
was achieved in ALL tacrolimus-treated dogs that
completed the trial, while the five that did not reach
this goal were those that failed to complete. The ABI
between tacrolimus and placebo group event (treatment success) rates was 0.75 (75%). Therefore, the
number of feet needed to be treated (NNT) with tacrolimus to achieve treatment success in one additional
case was two (1.3 rounded upwards to the next whole
number by convention).

Adverse drug events

At each re-evaluation visit, owners were asked to
describe whether they had observed any side effects

Figure 3. (a) Scatter plot of percentage change from baseline pedal

lesional scores during the clinical trial. Each point represents a study
subject. Closed circles are values from dogs that completed the trials.
Open circles correspond to missing values carried forward because
of withdrawal of study subjects. All data sets were kept for intentionto-treat analyses. (b) Column graph of percentage of study subjects
achieving treatment success, defined as the percentage of dogs exhibiting
a 50% reduction from baseline scores for the relevant intervention.

after applying either medication. Adverse effects were

not seen after the application of tacrolimus or placebo
in 15/20 dogs (75%). In five dogs (25%), the owners
reported occasional attempts by the dog to lick the site
of application of tacrolimus ointment for a few seconds
(two cases) to a few minutes (three cases) immediately after applying this drug. This licking behaviour
was interpreted subjectively as a sign of irritation.

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 52 60

Tacrolimus for localized atopic dermatitis

57

These reactions were mild and transient: they were

seen mainly only during the first few days of tacrolimus application, and did not necessitate the use
of any adjunctive therapy or cessation of ointment
application.

scores at least 50% lower than at baseline; this percentage reached 100% if only dogs that completed the study
were taken into account. A logical explanation for this
higher treatment effect might be the twice-daily application of tacrolimus ointment employed here compared
to the once-daily regimen selected in the other study.10

D ISCU SSION

Harm

Benefit
In this investigator-blinded placebo-controlled trial
enrolling 20 dogs, daily application of a commercially
available 0.1% tacrolimus ointment for 6 weeks was
shown to be effective for treatment of localized AD in
dogs. The topical application of tacrolimus ointment
resulted in a clinically relevant reduction of lesional
scores (63% median improvement from baseline
values) with 75% of dogs exhibiting treatment success,
defined as a greater than 50% reduction in scores on
tacrolimus-treated feet. Significantly, dogs that did not
reach this benchmark were those that did not complete
the study. In contrast to these positive results, the application of vaseline as a placebo did not lead to any
noticeable improvement in lesional scores (3% median
worsening from baseline values), and no dogs achieved
treatment success on vaseline-treated feet. As shown
with ABI and NNT calculations, only two dogs needed
to be medicated with tacrolimus to achieve one additional treatment success. Such a value is indicative of a
strong treatment effect.
The efficacy of tacrolimus for treatment of skin
lesions of canine AD appears to be remarkably similar
in magnitude to treatment in human atopic patients.7
Indeed, the percentage reduction from baseline
lesional scores varied between 40 and 77% in several
randomized controlled trials employing the 0.1%
tacrolimus ointment.1719
In veterinary dermatology, only two small-scale
RCTs have been carried out to evaluate the efficacy of
tacrolimus for treatment of canine AD.9,10 In the first
study, a compounded 0.3% tacrolimus lotion was
shown to be moderately effective to reduce erythema
(33% median reduction from baseline scores), but it
had inconsistent antipruritic effect depending on
whether owners or clinicians assessed this clinical
sign.9 The same authors conducted a second trial evaluating the efficacy of the commercially available 0.1%
tacrolimus ointment,10 a formulation identical to the
one tested in our study. In this second study, the perceived efficacy of tacrolimus ointment was higher than
with the compounded lotion, with 58% of tacrolimustreated dogs achieving a greater than 50% reduction
from baseline scores.10 Significantly, this benchmark
was reached more often by dogs with localized AD
(71%) than those with generalized skin lesions (24%).10
Our trial, which tested the efficacy of 0.1% tacrolimus
ointment on localized atopic skin lesions, yielded beneficial results superior than those obtained by Marsella
and colleagues.10 Indeed, using conservative ITT analyses,
75% of feet treated with tacrolimus for 6 weeks had

In human patients with AD, skin irritation and a burning sensation are the adverse drug events reported most
commonly, especially after treatment of head and neck
lesions.7,18,19 This side effect can be severe enough to
warrant discontinuation of drug application.18 In both
veterinary trials testing the efficacy of tacrolimus for
treatment of canine AD, the presence of inflammation
at the site of tacrolimus application was not specifically
discussed. In the present study, skin irritation a sensation interpreted herein by licking impulses of the site
of application was observed in five dogs (25%). In no
cases, however, was this side effect severe enough to
warrant withdrawal from the trial.
In both trials testing the efficacy of tacrolimus in
dogs with ,9,10 blood levels of the active medication
were measured, and they were shown to remain usually
below toxic levels. In these studies there were no
reported changes in blood counts and chemistry
parameters after tacrolimus administration.9,10 A limitation of the trial described herein is that blood levels
of tacrolimus, routine blood counts and chemistry panels
were not determined, but there were no systemic signs
suggesting impairment of function of any internal
organs. The impact of long-term application of tacrolimus
ointment on the health of dogs remains unknown.

Cost
The assessment of cost vs. benefit plays an important
role in therapeutic decision making. In this trial lasting
6 weeks, owners purchased only one 30-g tube of Protopic at the retail cost of 37.7 Euros (US$47). Whether
some medication remained in the tubes at the end of
this trial was not assessed. Therefore, a precise analysis
of treatment cost cannot be determined.
In the North American trial that evaluated the efficacy of tacrolimus ointment for treatment of canine
AD, a cost analysis was performed.10 The monthly cost
of treating a 25-kg dog with 0.1 mL kg1 of tacrolimus
once daily was estimated to be US$120 (100 Euros),
using one 60-g tube. This amount would be more than
adequate to treat localized skin lesions of AD, but it
would be insufficient for widespread lesions.
A recent study compared the cost and benefit of
treatment with tacrolimus ointment vs. high potency
topical glucocorticoids in human patients with moderate to severe AD unresponsive to mid-potency topical
glucocorticoids.20 It was found that, for monthly treatment regimens, high potency topical glucocorticoids
and tacrolimus ointments were similar in total costs
and efficacy. For 2-week treatments, tacrolimus was
found to be more cost effective than glucocorticoids.
Also, although the primary cost of tacrolimus ointment

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 5260

58

E Bensignor and T Olivry

was higher than that of glucocorticoids, patients

treated with high potency glucocorticoids incurred
higher secondary drug costs.20
In summary, the cost of tacrolimus ointment
remains high, but its remarkable efficacy is likely to be
cost effective, especially in dogs with localized or multifocal AD skin lesions.

Limitations
In this blinded RCT there are only a few limitations
that prevent the generalization of our findings to the
entire population of dogs with AD.
First, the outcome variable employed in this trial
relied on the use of a new scoring scale that has so far
not been evaluated for validity (content, construct and
criterion) and reliability as suggested by Charman and
Williams.21 It was felt, however, that the evaluation of
both a primary (erythema) and three secondary lesions
(lichenification, excoriations and oozing) was a realistic indication of the stage and severity of pedal signs of
canine AD. The grading of secondary lesions was
considered to represent a good marker of underlying
pruritus followed by secondary licking and chewing.
Moreover, this scoring of secondary skin lesions was
believed to compensate for the lack of formal evaluation of pruritus severity by either owner or clinician.
However, it is possible that, as shown previously9, tacrolimus ointment could have a stronger effect on reducing
skin lesions than on pruritus. Such variable effect could
decrease the perceived benefit of using tacrolimus ointment for treatment of canine AD.
Nevertheless, the evaluation of the combination of
lesions used was deemed to represent a good assessment
of the quality of life of study subjects. It was therefore
concluded that this appraisal of lesions would satisfy
the content parameter proposed by the authors cited
above.21 As only one clinician evaluated skin lesions
in all subjects at each re-evaluation visit, the need for
interobserver reliability validation was eliminated.
Second, only a small number of animals were treated
in this prospective trial, and this low number could
have altered the estimation of treatment effect. However, subjects enrolled were representative of the entire
atopic population, either in their selection from general and referral practice settings, overall breed representation, age of onset, or duration of disease prior to
enrolment, and the treatment effect recorded was very
large. Thus, the small scale of this trial is unlikely to
affect the application of its results to the canine atopic
population.
Third, we treated in this trial solely dogs with mild
to moderate AD, and we restricted our evaluation to
small lesional areas on the subjects feet. It is unknown
whether feet lesions would respond differently than
those at other sites. Furthermore, the usefulness of
tacrolimus for extensive and/or severe canine AD cannot be extrapolated from the present results.
Additionally, the exact amount of tacrolimus applied
onto skin lesions was not specified by the clinician to
the owner in the protocol employed. Consequently,

whether the efficacy of treatment would have been different if a lower or higher amount of medication had
been applied remains unknown. This trial only evaluated the efficacy of 0.1% tacrolimus ointment applied
twice daily for 6 weeks. As a result, this study has not
provided any information on the benefit of lower concentrations of tacrolimus (a 0.03% ointment is available commercially) or on the possibility of maintaining
a long-term beneficial effect with reduced frequency of
application of the ointment.
Finally, this trial has not provided any information
on whether tacrolimus ointment could offer additional
or synergistic benefit when given with oral medications
proven to be effective for treatment of AD in dogs.4 It
is indeed conceivable that the topical application of
tacrolimus ointment onto severe lesions at selected
body sites would permit a reduction of doses of oral
medications needed to maintain a good quality of life.
A reduction in overall treatment cost would be probable when using such combined intervention regimens.
There is a clear need for additional trials designed to
address the important clinical questions highlighted in
the preceding paragraphs.
In conclusion, results of this RCT together with those
of a previously performed trial10 provide evidence that
commercially available tacrolimus ointment is effective
for treatment of clinical signs of AD in dogs. Application of the ointment twice daily may be more beneficial
than once per day, but this assumption needs to be confirmed in a comparative trial. The long-term efficacy
and safety of tacrolimus ointment, especially with
decreasing application frequency regimens, is not
known. Harm appears to be minimal, except for transient behaviour suggesting irritation after application
during the first weeks of treatment. The results of both
trials, the nature of this formulation and the cost of the
drug suggest that tacrolimus ointment might be better
suited for treatment of localized than generalized AD
in dogs.

AC K N OW L E D G E M E N T S
The authors are grateful to Dr Malcolm Roberts from
NC State University for his review of the manuscript
and useful comments on the statistical analyses used.

REFERENCES
1. Olivry T, DeBoer DJ, Griffin CE et al. The ACVD task
force on canine atopic dermatitis: forewords and lexicon.
Veterinary Immunology and Immunopathology 2001;
81: 1436.
2. Hillier A, Griffin CE. The ACVD task force on canine
atopic dermatitis (I): incidence and prevalence. Veterinary
Immunology and Immunopathology 2001; 81: 14752.
3. Olivry T, Sousa CA. The ACVD task force on canine
atopic dermatitis (XIX): general principles of therapy.
Veterinary Immunology and Immunopathology 2001;
81: 31116.

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 52 60

Tacrolimus for localized atopic dermatitis

4. Olivry T, Mueller RS. Evidence-based veterinary dermatology: a systematic review of the pharmacotherapy of
canine atopic dermatitis. Veterinary Dermatology 2003;
14: 121 46.
5. Bekersky I, Fitzsimmons W, Tanase A et al. Nonclinical
and early clinical development of tacrolimus ointment
for the treatment of atopic dermatitis. Journal of the
American Academy of Dermatology 2001; 44: S1727.
6. Nghiem P, Pearson G, Langley RG. Tacrolimus and
pimecrolimus: from clever prokaryotes to inhibiting calcineurin and treating atopic dermatitis. Journal of the
American Academy of Dermatology 2002; 46: 22841.
7. Williams H, Thomas K, Smethurst D et al. In: Williams
HC, Bigby M, Diepgen T et al. eds. Evidence-based Dermatology. London, UK: BMJ Books, 2003: 144218.
8. Yetman RJ, Parks D. Diagnosis and management of
atopic dermatitis. Journal of Pediatric Healthcare 2002;
16: 143 5.
9. Marsella R, Nicklin CF. Investigation on the use of 0.3%
tacrolimus lotion for canine atopic dermatitis: a pilot
study. Veterinary Dermatology 2002; 13: 203 10.
10. Marsella R, Nicklin CF, Saglio S et al. Investigation
of the clinical efficacy and safety of 0.1% tacrolimus
ointment (Protopic) in canine atopic dermatitis: a
randomized, double blinded, placebo controlled, crossover study. Veterinary Dermatology 2004; 15: 294303.
11. Willemse T. Atopic skin disease: a review and a reconsideration of diagnostic criteria. Journal of Small Animal
Practice 1986; 27: 771 8.
12. Prlaud P, Guagure E, Alhaidari Z et al. [Reevaluation
of diagnostic criteria of canine atopic dermatitis]. Revue
de Medecine Veterinaire 1998; 149: 1057 64.
13. DeBoer DJ, Hillier A. The ACVD task force on canine
atopic dermatitis (XV): fundamental concepts in clinical

14.

15.

16.
17.

18.

19.

20.

21.

59

diagnosis. Veterinary Immunology and Immunopathology 2001; 81: 2716.

Olivry T, Rivierre C, Jackson HA et al. Cyclosporine
decreases skin lesions and pruritus in dogs with atopic
dermatitis: a blinded randomized prednisolone-controlled
trial. Veterinary Dermatology 2002; 13: 7787.
Olivry T, Steffan J, Fisch RD et al. Randomized controlled trial of the efficacy of cyclosporine in the treatment of
atopic dermatitis in dogs. Journal of the American Veterinary Medical Association 2002; 221: 3707.
Cockcroft P, Holmes M. Evidence-Based Veterinary
Medicine. Oxford, UK: Blackwell Publishing.
Boguniewicz M, Fiedler VC, Raimer S et al. A randomized, vehicle-controlled trial of tacrolimus ointment
for treatment of atopic dermatitis in children. Journal of
Allergy and Clinical Immunology 1998; 102: 63744.
Reitamo S, Rustin M, Ruzicka T et al. Efficacy and
safety of tacrolimus ointment compared with that of
hydrocortisone butyrate ointment in adult patients with
atopic dermatitis. Journal of Allergy and Clinical Immunology 2002; 109: 54755.
Reitamo S, Van Leent EJ, Ho V et al. Efficacy and safety
of tacrolimus ointment compared with that of hydrocortisone acetate ointment in children with atopic dermatitis. Journal of Allergy and Clinical Immunology
2002; 109: 53946.
Ellis CN, Drake LA, Prendergast MM et al. Costeffectiveness analysis of tacrolimus ointment versus highpotency topical corticosteroids in adults with moderate
to severe atopic dermatitis. Journal of the American
Academy of Dermatology 2003; 48: 55363.
Charman C, Williams H. Outcome measures of disease
severity in atopic eczema. Archives of Dermatology
2000; 136: 7639.

Rsum Cette tude randomise, contrle, en aveugle, a t mise en place pour dterminer si une pommade
au tacrolimus (Protopic, Fujisawa) diminuait la svrit de lsions localises de dermatite atopique canine (AD).
Vingt chiens prsentant une AD ont t inclus, sils prsentaient des lsions des deux membres antrieurs. Chaque
pied a t trait soir avec le tacrolimus soit avec le placebo (vaseline) deux fois par jour pendant 6 semaines. Avant
et toutes les deux semaines pendant ltude, lrythme, la lichnification, les suintements et lexcoriation ont t
grads sur une chelle allant de 0 10 (score maximum de 40). Les critres tudis taient le pourcentage de rduction des scores lsionnels depuis linclusion, et le nombre de sujets ayant un score diminu de plus de 50% la
fin de lessai. Les analyses ont t faites selon la technique intention to treat. Au dbut de lessai, les scores
lsionnels ntaient pas diffrents entre les zones traites par le tacrolimus ou le placebo. Aprs six semaines, le
pourcentage de rduction lsionnelle tait plus lev pour les zones lses traites par le tacrolimus (moyenne:
63%; 95% intervalle de confiance: 39 67]) que pour les pieds traits par le placebo (moyenne: 3%; intervalle de
confiance: 2 13) (Wilcoxon test; P = 0.0003). Aprs traitement par le tacrolimus, les lsions ont diminu de
50% ou plus chez 15 des 20 chiens (75%); ces animaux ont termin ltude. A loppos, cette limite na t atteinte
pour aucun des pieds traits par le placebo (Fishers test; P < 0.0001). Des ractions secondaires minimes ont
t notes: irritation dans certaines zones lses traites par le tacrolimus. Les rsultats de cette tude suggrent
que lapplication dune pommade 0.1% de tacrolimus est utile pour diminuer la gravit des lsions localises
de dermatite atopique canine.
Resumen Esta prueba controlada al azar con investigadores que examinaban a ciegas, fue diseada para determinar si la pomada de tacrolimus (Protopic, Fujisawa) disminua la gravedad de las lesiones locales de la dermatitis atpica canina (AD). Se seleccionaron veinte perros con AD si mostraban lesiones en ambos metacarpos
frontales. Cada extremidad se trat, al azar, con pomada de 0.1% tacrolimus o con placebo (vaselina) dos veces
al da durante seis semanas. Anteriormente, y cada dos semanas durante el estudio, se puntu con una escala
de 10 puntos la intensidad del eritema, la liquenificacin, la supuracin y las excoriaciones (puntuacin mxima
total: 40). Las mediciones de los primeros resultados fueron el porcentaje de reduccin a partir de los valores
de las lesiones iniciales y el nmero de individuos cuyos valores haban disminuido un 50% o ms al final del estudio. Se utilizaron anlisis segn la intencin de tratar. Al inicio del estudio, las puntuaciones de las lesiones no
2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 5260

60

E Bensignor and T Olivry

eran significativamente diferentes entre reas tratadas con tacrolimus o placebo. A las seis semanas, el porcentaje
de reduccin a partir de los valores iniciales era superior en las reas tratadas con tacrolimus (media: 63%; 95%
intervalo de confianza: 39 a 67]) que en las extremidades tratadas con placebo (media: 3%; intervalo de confianza:
2 a 13) (test de Wilcoxon; P = 0.0003). Cuando se aplicaba el tacrolimus, las lesiones disminuan en un 50% o
ms en 15 de 20 perros (75%); estos perros eran los que completaron el estudio. En contraste, esta cota no fue
alcanzada por ninguna de las extremidades tratadas con placebo (test de Fishers; P < 0.0001). Se detectaron
hallazgos adversos medicamentosos consistentes en irritaciones menores en algunas reas lesionales tratadas con
tacrolimus. Los resultados de esta prueba sugieren que la aplicacin de pomada de tacrolimus al 0.1% resulta
til para reducir la gravedad de las lesiones cutneas localizadas en la AD canina.

Zusammenfassung Diese Untersucher-verblindete, randomisierte, kontrollierte Untersuchung wurde

konzipiert, um zu bestimmen, ob Tacrolimus-Salbe (Protopic, Fujisawa) den Schweregrad lokalisierter Lsionen
von caniner atopischer Dermatitis (AD) verringert. Einundzwanzig Hunde mit AD wurden in die Studie aufgenommen, wenn sie Lsionen an beiden Metacarpi aufwiesen. Jeder Fu wurde nach dem Zufallsprinzip
entweder mit 0,1% Tacrolimus oder mit Placebo (Vaseline) zweimal tglich ber sechs Wochen behandelt. Vor
Beginn und alle 2 Wochen whrend der Studie wurden Erythem, Lichenifikation, Exsudation und Exkoriationen
mit einer 10-Punkte-Skala (maximale Gesamtbewertung:40) bewertet. Die primren Parameter fr das Ergebnis
waren prozentuale Reduktion von usprnglicher Punktzahl fr Lsionen und Anzahl der Individuen, deren
Punktzahl sich bei Studienende um 50% oder mehr verringert hat. Es wurde Intention-to-treat-Analyse angewandt. Zu Beginn der Studie war die Punktzahl fr Lsionen an den mit Tacrolimus oder Placebo behandelten
Lokalisationen nicht signifikant unterschiedlich. Nach sechs Wochen war die prozentuale Reduktion der
usprnglichen Punktzahl fr Lsionen bei den mit Tacrolimus behandelten Lokalisationen hher (durchschnittlich: 63%; 95% Konfidenzintervall: 39 to 67]) als bei den Placebo-behandelten Fen (durchschnittlich: 3%; Konfidenzintervall: 2 to 13) (Wilcox Test; P = 0.0003). Nach Tacrolimusapplikation verringerten sich die Lsionen
bei 15 von 20 Hunden (75%) um 50% oder mehr; diese Hunde waren diejenigen, die die Studie beendet haben.
Im Gegensatz hierzu wurde dieser Bezugswert bei keinem der Placebo-behandelten Fe erreicht (Fisher-Test;
P < 0.0001). Unvertrglichkeitsreaktionen nach Behandlung mit Tacrolimus waren geringgradige Irritationen in
einigen vernderten Bereichen. Die Ergebnisse dieser Studie lassen vermuten, da die Applikation von 0,1%-iger
Tacrolimussalbe zur Reduktion des Schweregrades lokalisierter Hautlsionen aufgrund von caniner AD ntzlich ist.

2005 European Society of Veterinary Dermatology, Veterinary Dermatology, 16, 52 60