EVIDENCE-BASED REVIEW

Treatments for Cutaneous Lichen Planus: A Systematic Reviewand Meta-Analysis

Lihi Atzmony1 • Ofer Reiter1 • Emmilia Hodak1,2 • Michael Gdalevich3 • Daniel Mimouni1,2

� Springer International Publishing Switzerland 2015

Abstract

Background Cutaneous lichen planus (CLP) is an

inflammatory dermatosis. Its chronic relapsing course and

frequently spontaneous regression hamper the assessment

of treatment effectiveness.

Objective To evaluate the efficacy of available treatment

modalities for CLP.

Data Sources PubMed, Cochrane Central Register of

Controlled Trials (CENTRAL), ClinicalTrials.gov registry.

Methods We performed a systematic review of the cur-

rent literature. All randomized controlled trials, nonran-

domized case–control studies, and cohort studies with more

than one treatment arm were included. The primary out-

comes were complete response and time to complete

response. The secondary outcomes were partial response,

relapse, time to relapse, reduction of itch, the adverse event

rate, and withdrawal due to adverse events.

Data Synthesis Sixteen studies met the inclusion criteria,

of which 11 were randomized controlled trials. Most trials

had a small sample size. In the rare studies in which

variants other than generalized or classic lichen planus

were included, they could not be analyzed separately.

Body-of-evidence quality ranged from very low to mod-

erate. Acitretin, sulfasalazine, and griseofulvin were

associated with increased overall response rates in com-

parison with placebo. Narrow-band ultraviolet B radiation

(NBUVB) was more effective than 6 weeks’ low-dose

prednisolone in achieving a complete response, and

prednisolone was more effective than enoxaparin.

Hydroxychloroquine was more effective than griseofulvin

in achieving an overall response. Betamethasone valerate

0.1 % ointment had comparable efficacy to calcipotriol

ointment. Methotrexate was effective, with a nonsignifi-

cant difference in the complete response rate in compar-

ison with oral betamethasone. In nonrandomized

controlled trials, oral psoralen plus ultraviolet A pho-

tochemotherapy (PUVA) had comparable efficacy to a

PUVA bath and NBUVB. Psoralen plus sunlight exposure

(PUVASOL) and betamethasone dipropionate 0.05 %

cream were effective relative to a short course of oral

metronidazole.

Conclusions Several effective treatment options are

available for CLP. Further well-designed studies are war-

ranted to investigate the efficacy of topical glucocorti-

coids—the current first-line therapy—as well as other

treatment modalities, and the treatment of different variants

of CLP.

L. Atzmony and O. Reiter contributed equally to this work.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s40257-015-0160-6) contains supplementarymaterial, which is available to authorized users.

& Daniel Mimouni

mimouni@post.tau.ac.il

1 Department of Dermatology, Rabin Medical Center,

Beilinson Hospital, 39 Jabotinski Street, 49100 Petach Tikva,

Israel

2 Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv,

Israel

3 Faculty of Health Sciences, Ben-Gurion University of the

Negev and South District Health Office, Ministry of Health,

Beer Sheva, Israel

Am J Clin Dermatol

DOI 10.1007/s40257-015-0160-6

Key Points

All treatments for cutaneous lichen planus that were

tested in comparative studies (randomized controlled

trials, nonrandomized controlled trials, and cohorts

with more than one treatment arm) were evaluated in

this systematic review.

Several treatment modalities were found to be

effective in randomized controlled trials, including

acitretin, sulfasalazine, narrow-band ultraviolet B

radiation, oral glucocorticoids, and griseofulvin. The

results are mainly applicable to classic or generalized

cutaneous lichen planus.

Further well-designed studies are warranted and

should focus on both classic cutaneous lichen planus

and other variants of this disease.

1 Introduction

Lichen planus is a common idiopathic inflammatory dis-

ease, affecting the skin, mucous membranes, hair, and

nails. Specifically, cutaneous lichen planus (CLP) affects

0.2–1 % of the adult population [1]. However, its treatment

is based mainly on the results of small studies and anec-

dotal reports, and remains controversial.

In the last 30 years, three systematic reviews on the

treatment of CLP have been reported. The first, which

appeared in 1996, included three randomized controlled

trials (RCTs) [2]; the second systematic review, from

2014, included eight RCTs and one additional study of

the same patient population, which was counted as a

separate trial [3]; and the third, also from 2014, included

only three RCTs [4]. None of the reviews was conducted

in accordance with the Preferred Reporting Items for

Systematic Reviews and Meta-Analyses (PRISMA)

statement [5].

To assess the efficacy of various treatment modalities

for CLP, we performed a systematic review and meta-

analysis of all relevant RCTs, case–control studies, and

cohort studies with two or more treatment arms.

2 Methods

This systematic review was conducted in accordance with

the PRISMA statement [5].

2.1 Eligibility Criteria

Only studies that met the following criteria were included

in the analysis:

• Language: English

• Study design: RCTs, quasi-randomized and nonran-

domized controlled trials, and cohort studies with more

than one treatment arm

• Population: patients of any age with CLP, with or

without mucosal lichen planus, who were diagnosed at

least according to clinical features

2.2 Outcomes

2.2.1 Primary Outcomes

The primary outcomes of the analysis were as follows:

• Proportion of patients achieving a complete response:

complete response was defined as disappearance of at

least 80 % of active cutaneous lesions. Post-inflammatory

hyperpigmentation was considered inactive disease.

• Time to complete response.

2.2.2 Secondary Outcomes

The first secondary outcome was the proportion of patients

achieving a partial response. Partial response was defined

as disappearance of 50–79 % of active cutaneous lesions.

Accordingly, overall response was defined as complete

response plus partial response.

If the above primary and/or secondary outcomes were not

reported, other endpoints with different definition of response

were considered. Additional secondary outcomes were:

• Proportion of patients with a relapse: relapse was

defined as the recurrence of active cutaneous lesions

after initial improvement.

• Time to relapse.

• Reduction of itch: itch was rated by a visual analog

scale or another scale detailed in the study.

• Rate of adverse events.

• Rate of withdrawal due to adverse events.

2.3 Literature Search

One reviewer (O.R.) searched the Cochrane Central

Register of Controlled Trials (CENTRAL; from inception

until May 2014), using the term ‘‘lichen planus’’, and

searched PubMed, using the Medical Subject Heading

(MeSH) term ‘‘lichen planus’’ and free-text words. The

L. Atzmony et al.

PubMed search strategy is detailed in Electronic Supple-

mentary Material 1. In addition, the ongoing trials registry

of the US National Institutes of Health (http://www.

clinicaltrials.gov) was screened for additional trials that

published results. The reference lists of the included trials

were searched for relevant publications. Authors were

contacted for missing data and clarifications.

2.4 Study Selection and Data Extraction

One reviewer (O. R.) screened the titles and abstracts of all

retrieved articles. Subsequently, when the titles and/or

abstracts suggested potential eligibility for the review, the

same author screened the full texts and extracted the rel-

evant data into a predefined electronic form. A second

reviewer (L.A.) checked the extraction of the data from all

included studies. Disagreements were resolved by discus-

sion with a third reviewer (D. M.).

2.5 Risk-of-Bias Assessment

The risk of bias was assessed in both randomized and

nonrandomized studies, using the criteria generated by

Downs and Black [6]. The checklist included 27 items

divided into five subcategories. The maximum achievable

scores for each of the subcategories were as follows: 10 for

reporting; 3 for external validity; 7 for internal validity—

bias in the measurement of the intervention and outcomes;

6 for internal validity—confounding (selection bias); and

1 for power. The total achievable maximum score was 27.

For the present review, the power item was simplified to a

simple check of whether the study included a statistical

power calculation. Downs and Black score ranges were

grouped into the following four quality levels: excellent

(26–27); good (20–25); fair (14–19); and poor (B13).

2.6 Strength of Body of Evidence

The quality of the evidence was graded as high, moderate,

low, or very low, as defined by the Grading of Recom-

mendations Assessment, Development and Evaluation

(GRADE) system [7].

2.7 Data Analysis and Synthesis

We calculated the relative risks (RRs) and 95 % confidence

intervals (CIs). Intention-to-treat (ITT) analysis was

applied when the data were adequate, and available case

analysis was used when they were not. When possible,

qualitative analysis using Review Manager (RevMan)

version 5.3 was performed with the random-effects model.

Otherwise, a qualitative synthesis was done.

The heterogeneity of the trial results was assessed by

visual examination of the forest plot to detect non-over-

lapping CIs, using a chi-squared test of heterogeneity (with

P\ 0.1 indicating statistical significance) and an I2

statistic for inconsistency (with 30–60 % denoting mod-

erate levels of heterogeneity).

3 Results

Our search yielded 2040 studies (Fig. 1). Sixteen studies

fulfilled the eligibility criteria, including 11 RCTs [8–18],

one adverse event report of an already included trial com-

bined with the initial trial data [19], three nonrandomized

prospective case–control studies [20–22], and one nonran-

domized retrospective case–control study [23]. The char-

acteristics of the included trials are detailed in Table 1. The

available data were not adequate to clearly support one

treatment approach over another. Most of the RCTs that

were included were rated as fair, using the Downs and Black

checklist (Table 2). The detailed appraisal of risk of bias is

shown in Electronic Supplementary Material 2. Six trials

included patients with generalized CLP [9, 12, 14, 15, 19,

20]. None of the trials reported treatment for specific CLP

variants separately (such as lichen planopilaris or lichen

planus pigmentosus). Only two trials included a few patients

with hypertrophic CLP—a much more resistant subtype—

precluding analysis of this subgroup separately [20, 21].

The definitions of the outcome measures varied among

the studies, as did the timing of their appraisal, which

ranged from 4 to 12 weeks; in one trial [12], the timing was

not stated. Most studies reported dichotomous clinical

response data. Time-to-event outcomes either were not

reported or were reported using inappropriate statistical

methods in all included trials but one [23], and so they

were not included in this review.

3.1 Phototherapy

Four of the included studies investigated the efficacy of

phototherapy in CLP [13, 20, 21, 23]. In the only RCT,

52.17 % of patients (12/23) treated with narrow-band

ultraviolet B radiation (NBUVB) achieved a complete

response, compared with 13.04 % of patients (3/12) treated

with prednisolone 0.3 mg/kg/day for 6 weeks (RR 4

[95 % CI 1.3–12.33]) [13].

The second study—a retrospective cohort study—com-

pared NBUVB with oral psoralen plus ultraviolet A pho-

tochemotherapy (PUVA) [23]. The complete response rate

was higher for PUVA, but the difference did not reach

statistical significance (4/13 versus 10/15, RR 0.46

[95 % CI 0.19–1.12]). There were no between-group dif-

ferences in the partial response rate (6/13 versus 5/15,

Treatments for Cutaneous Lichen Planus

RR 1.38 [95 % CI 0.55–3.49]) and the overall response

rate (77 versus 100 %). It is of note that two patients with

hypertrophic lichen planus lesions (one in each interven-

tion group) did not respond to NBUVB and had a partial

response to PUVA. Another two patients who did not

respond to NBUVB were switched to PUVA and thereafter

achieved a complete response. Relapse rates for the two

treatments were comparable (3/10 versus 7/15, RR 0.64

[95 % CI 0.55–3.49]), as were the times to relapse (rates of

sustained response at 1 month follow-up: 72 % for PUVA

versus 54 % for NBUVB, P = 0.86) and the rates of

adverse events (RR 0.23 [95 % CI 0.01–4.37]).

The third study—a non-RCT—compared the efficacy of

oral PUVA with a PUVA bath [20]. The overall response

rate was nonsignificantly higher for the PUVA bath (11/13

versus 5/10, RR 1.69 [95 % CI 0.87–3.28]). The non-

significant difference was attributable mostly to the partial

response rate (9/13 versus 4/10), not the complete response

rate (2/13 versus 1/10).

The fourth study—a non-RCT—compared oral

8-methoxypsoralen plus sunlight exposure (PUVASOL)

with oral metronidazole 200 mg three times daily and with

cetirizine plus betamethasone dipropionate 0.05 % cream

[21]. The rate of overall ‘‘clinical improvement’’ (from a

mild decrease in symptoms and in lesion size and number

to complete disappearance of symptoms and lesions) was

significantly higher for PUVASOL (70.8 %) and for ceti-

rizine plus betamethasone dipropionate (69.6 %) than for

metronidazole (34.8 %, RR 1.92 [95 % CI 1.02–3.59] and

RR 2.04 [95 % CI 1.1–3.77], respectively). However,

metronidazole was administered for only 3 weeks, whereas

the other treatments were administered for 12 weeks. In

addition, the definition of ‘‘response’’ was very permissive.

There was no between-group difference in the rate of

adverse events.

3.2 Systemic Therapy

3.2.1 Oral Glucocorticoids

Three RCTs compared the efficacy of oral glucocorticoids

with other therapies including enoxaparin, NBUVB, and

Fig. 1 Study flow chart. CENTRAL Cochrane Central Register of Controlled Trials

L. Atzmony et al.

Table

1Characteristicsoftrials

included

inthereview

Trial

IDCountry

Studydesign

Drugarms

Regim

ens

Assessm

ent

time

NAge(years)

Fem

ale

sex(%

)

CLPtypes

Follow-up

Bhuiyan,

2010[12]

Bangladesh

RCT

PO hydroxychloroquine

400mg/day,6months

NS

80

Range:

20–60

33.75

‘‘Classic’’

1.5

years

PO

griseofulvin

500mg/day,6months

Bouloc,

2000[8]

France

RCT

KH

1060

BID

,8weeks

12weeks

92

Mean:50.2

NS

NS

12weeks

Vehicle

BID

,8weeks

Chopra,

1999[22]

India

Non-RCT

Topical

betam

ethasone

0.1

%BID

,3months

1–3months

75

60%

30–60years

old

48

LPvulgaris,

hypertrophic,

actinicus,linearis,

follicular,atrophicus,

annularis,

pigmentosus,mucous

mem

brane

3months

PO

dapsone

50mgTID

,3months

Hazra,

2013[9,19]

Bangladesh

RCT

PO

methotrexate

10mgonce

weekly,

12weeks

12weeks

44

Mean:33.9

63.6

Generalized

12weeks

PO

betam

ethasone

5mgtwiceweekly,12weeks

Helander,

1987[20]

Finland

Non-RCT

PO

methoxalen

0.6

mg/kg?

UVA

Range:

8–46exposures

EndofPUVA

treatm

ent

99

Mean:43

NS

Generalized

?localized

Range:

10–43months

50mgmethoxalen

bath?

UVA

Notreatm

ent

Iraji,

2011[13]

Iran

RCT

NBUVB

3times

weekly,70%

of

MED,6weeks

6weeks

46

Mean±

SD:

prednisone

42.04±

2.46;

NBUVB

36.13±

2.88

74

Generalized

6weeks

PO

prednisolone

0.3

mg/kg/day,6weeks

Iraji,

2013[14]

Iran

RCT

SCenoxaparin

5mgonce

weekly,

max

8weeks

8weeks

54

Mean:enoxaparin

38.8;

prednisone36.7

60.4

Generalized

6months

PO

prednisone

0.5

mg/kg/day,max

8weeks

Kellett,

1990[10]

UK

RCT

PO

prednisone

30mg/day

?hydrocortisone

17butyrate

cream

BID

,

10days

6weeks

30

NS

NS

NS

2years

PO

placebo

OD

?hydrocortisone

17butyrate

cream

BID

,

10days

Laurberg,

1991[11]

Denmark

RCT

PO

acitretin

30mg/day,8weeks

8weeks

65

Mean:acitretin

44.5;placebo

53

49.2

NS

16weeks

PO

placebo

OD,8weeks

Omidian,

2010[15]

Iran

RCT

PO

sulfasalazine

2.5

g/day,6weeks

6weeks

52

Mean:33.45

NS

Generalized

6weeks

PO

placebo

OD,6weeks

Treatments for Cutaneous Lichen Planus

Table

1continued

Trial

IDCountry

Studydesign

Drugarms

Regim

ens

Assessm

ent

time

NAge(years)

Fem

ale

sex(%

)

CLPtypes

Follow-up

Sehgal,

1972[17]

India

RCT

PO

griseofulvin

500mgOD,4weeks

4–6weeks

24

Range:

10–55

55.9

NS

4–6weeks

PO

placebo

OD,4weeks

Sehgal,

1980[18]

India

RCT

PO

griseofulvin

500mgOD,8weeks

8weeks

44

Range:

8–70

45.4

NS

8weeks

PO

placebo

OD,8weeks

Sharma,

2003[21]

India

Non-RCT

PO

metronidazole

200mgTID

,3weeks

12weeks

70

Mean:33.71

51.4

Classic,hypertrophic,

guttate,

atrophic

NS

PO

oxoralen

4mg/kg?

30min

sunlight

3times

weekly,12weeks

PO

cetirizine

10mg?

topical

betam

ethasonedipropionate

cream

0.05%

OD,

12weeks

Theng,

2004[16]

Singapore

RCT

Calcipotriolointm

ent

50mg/g

BID

,12weeks

12weeks

31

Range:

14–75

52

Allbut1patienthad

multiple-site

involvem

ent(lim

bs

andtrunk)

12weeks

Betam

ethasone

valerateointm

ent

0.1

%BID

,12weeks

Wackernagel,

2007[23]

Austria

Retrospective

cohort

PO

PUVA

PO

oxoralen0.6

mg/

kg?

UVA;mean:25.9

exposures(range:

9–34)

Mean:

10.5

weeks

(range:

4–18)

Mean:

8.2

weeks

(range:

5–12)

18

Mean:PUVA

47

(range:

16–65);

NBUVB51

(range:

19–69)

75

Generalized;1patientin

each

grouphad

hypertrophiclesionsin

addition

Mean:

20.5

months

(range:

2–49)

NBUVB

Mean:22.5

exposures(range:

15–31)

Mean:

35.7

months

(range:

3–60)

BID

twotimes

daily,CLPcutaneouslichen

planus,

LPlichen

planus,

maxmaxim

um,MED

minim

alerythem

atousdose,NBUVBnarrow-bandultravioletB

radiation,NSnotstated,OD

once

daily,

PO

per

os,PUVApsoralen?

ultravioletA

photochem

otherapy,RCTrandomized

controlled

trial,SCsubcutaneous,TID

threetimes

daily,UVAultravioletA

L. Atzmony et al.

methotrexate [9, 13, 14], and one RCT compared the

efficacy of oral prednisolone with placebo [10]. Overall,

the complete response rates ranged from 13.04 to 59.26 %.

Specifically, in the study of 10 days’ prednisolone

(30 mg/day) versus placebo, the complete response rates at

6 weeks were comparable in the two arms (4/14 versus

Table 2 Efficacy outcomes of trials included in the review

Trial ID Drug arms (N) Overall

response,

% (n)

Partial

response,

% (n)

Complete

response,

% (n)

Risk-of-bias

ratingdITTe/per-

protocol

analysis

Bhuiyan, 2010 [12] PO hydroxychloroquine (N = 40) 70 (28)* 52.5 (21) 17.5 (7) 18 NS

PO griseofulvin (N = 40) 42.5 (17)* 37.5 (15) 5 (2)

Bouloc, 2000 [8] KH 1060 (N = 49) NS 28.6 (14) NS 19 ITT

Vehicle (N = 43) 34.8 (15)

Chopra, 1999 [22]a Topical betamethasone 0.1 % (N = 25) 40 (10) NS NS 8 NS

PO dapsone (N = 59) 49.2 (29)

Hazra, 2013 [9] PO methotrexate (N = 23) NS NS 69.6 (16) 18 Per-protocol

PO betamethasone (N = 21) 47.6 (10)

Helander, 1987 [20] PO PUVA (N = 10) 50 (5) 40 (4) 10 (1) 9 Per-protocol

PUVA bath (N = 13) 84.6 (11) 69.2 (9) 15.4 (2)

Iraji, 2011 [13] NBUVB (N = 23) NS NS 52.2 (12)* 16 ITT

PO prednisolone (N = 12) 13 (3)*

Iraji, 2013 [14] SC enoxaparin (N = 27) 66.7 (18) 37 (10) 29.6 (8)* 20 ITT

PO prednisone (N = 27) 81.5 (22) 22.2 (6) 59.3 (16)*

Kellet, 1990 [10] PO prednisone (N = 14) NS NS 28.6 (4) 14 ITT

PO placebo (N = 14) 7.1 (1)

Laurberg, 1991 [11]a PO acitretin (N = 32) 56.3 (18)* 37.5 (12)* 18.75 (6) 22 ITT

PO placebo (N = 33) 12.1 (4)* 9 (3)* 3 (1)

Omidian, 2010 [15] PO sulfasalazine (N = 26) 73 (19)* 42.3 (11)* 30.8 (8) 21 ITT

PO placebo (N = 26) 7.7 (2)* 3.8 (1)* 3.8 (1)

Sehgal, 1972 [17] PO griseofulvin (N = 17) NS NS 70.6 (12) 14 Per-protocol

PO placebo (N = 17) 35.3 (6)

Sehgal, 1980 [18] PO griseofulvin (N = 22) 100 (22)* 18.2 (4)a 81.8 (18)* 16 Per-protocol

PO placebo (N = 22) 22.7 (5)* 22.7 (5)a 0*

Sharma, 2003 [21] PO metronidazole (N = 23) 34.8 (8)*,b NS NS 11 Per-protocol

PUVASOL (N = 23) 70.8 (16)*,b

PO cetirizine 10 mg ? betamethasone

dipropionate 0.05 % cream (N = 24)

69.6 (17)*,b

Theng, 2004 [16] Calcipotriol 0.05 % ointment (N = 15) 47.6 (7)c 6.6 (1) NS 19 ITT

Betamethasone valerate 0.1 % ointment

(N = 16)

50 (8)c 0 NS

Wackernagel, 2007 [23] PO PUVA (N = 15) 100 (15) 33.3 (5) 66.7 (10) 15 Per-protocol

NBUVB (N = 13) 77 (10) 46.2 (6) 30.8 (4)

ITT intention-to-treat, NBUVB narrow-band ultraviolet B radiation, NS not stated, PO per os, PUVA psoralen ? ultraviolet A pho-

tochemotherapy, PUVASOL psoralen ? sunlight exposure, SC subcutaneous

* A statistically significant difference was observed (P\ 0.05)a A detailed definition of ‘‘response’’ was not supplied in the original articleb Overall response = clinical improvement defined as a range of responses from a mild decrease in symptoms and in lesion size and number to

complete disappearance of symptoms and lesions; clinical improvement was significantly greater with PUVASOL or betamethasone than with

metronidazolec Overall response = lesion flatteningd The risk of bias was assessed using the Downs and Black checklist [6]; the total achievable maximum score was 27e ITT/modified-ITT

Treatments for Cutaneous Lichen Planus

1/14, RR 4 [95 % CI 0.51–31.46]), although improvement,

graded by a linear analog severity scale, was significantly

greater in the prednisolone group (the median score at entry

minus the score at 6 weeks: 2.5 versus 0.85, P\ 0.05).

On meta-analysis of the complete response rate with oral

glucocorticoids against other therapies, the pooled RR was

0.76 [95 % CI 0.27–2.09], with substantial heterogeneity

(Fig. 2). Because of the small number of included trials, the

heterogeneity could not be resolved with subgroup or

sensitivity analysis. However, when each trial was exam-

ined separately, the complete response rate was signifi-

cantly higher for oral prednisone than for enoxaparin [14],

and lower than for NBUVB [13]. There was no significant

difference in the complete response rate between oral

betamethasone 5 mg twice weekly and methotrexate

10 mg/week [9]. In the only RCT (prednisone versus

enoxaparin) reporting the partial response rate [14], no

between-group difference was shown. Two trials reported

the relapse rate [10, 14]: one reported a nonsignificantly

higher rate for prednisone than for placebo (6/16 versus

1/14, RR 4.38 [95 % CI 0.58–33.1]), and the other found

no difference between prednisone and enoxaparin (6/18

versus 9/22, RR 0.81 [95 % CI 0.36–1.86]). Three trials

reported adverse events [10, 14, 19], but their method of

reporting varied. A nonsignificantly higher proportion of

patients experienced adverse events with prednisone than

with enoxaparin or placebo (5/27 versus 1/27 and 2/16

versus 0/14, respectively). After 12 weeks of treatment,

prednisone was associated with more events of anemia,

edema, dyspepsia, acne, moon face, striae, and menstrual

abnormalities than methotrexate, but there were no

between-group differences in blood cell counts, liver

function tests, and plasma glucose levels (P\ 0.05).

3.2.2 Acitretin

One RCT compared the efficacy of 30 mg acitretin daily

for 8 weeks with placebo [11]. The overall response rate

was higher for acitretin (18/32 versus 4/33, RR 4.64

[95 % CI 1.76–12.22]), with no between-group difference

in the complete response rate (6/32 versus 1/32, RR 6.19

[95 % CI 0.79–48.57]). In addition, the acitretin group had

a significantly higher proportion of patients with pruritus at

baseline and a significantly lower proportion after 8 weeks

of treatment (50 versus 75.8 %, RR 0.66 [95 % CI

0.44–0.98]). It is worth noting that adverse events occurred

at a significantly higher rate in the acitretin group (88

versus 52 %, RR 1.7 [95 % CI 1.19–2.42]), but none of the

patients withdrew from the study for this reason.

3.2.3 Griseofulvin

Three trials assessed the efficacy of griseofulvin [12, 17,

18]—two against placebo [17, 18] and one against

hydroxychloroquine [12]. In comparison with placebo, 4 or

8 weeks’ treatment with griseofulvin 500 mg/day was

associated with a higher overall response rate (34/39 versus

11/39, RR 2.85 [95 % CI 1.4–5.79]), with moderate

heterogeneity (I2 = 48 %), and a nonsignificantly higher

complete response rate (30/39 versus 6/39, RR 7.28

[95 % CI 0.15–361.6]), with substantial heterogeneity

(I2 = 87 %). One trial that reported the relapse rate found

no cases of relapse during 9 months of follow-up [17]. In

comparison with hydroxychloroquine 400 mg/day, how-

ever, 6 months’ treatment with griseofulvin 500 mg/day

was associated with a significantly lower response rate

(42.5 versus 70 %, RR 0.6 [95 % CI 0.4–0.92]) [12]. There

were no between-group differences in the partial response

and complete response rates (RR 0.71 [95 % CI 0.43–1.17]

and RR 0.29 [95 % CI 0.06–1.29], respectively). Impor-

tantly, the timing of the outcome appraisal was not stated.

3.2.4 Sulfasalazine

Sulfasalazine 2.5 mg/day for 6 weeks was compared with

placebo in one trial [15]. There was no between-group differ-

ence in the complete response rate (8/26 versus 1/26, RR 8

[95 % CI 1.08–59.5]), although sulfasalazine was associated

with significantly higher partial response (11/26 versus 1/26,

RR 11 [95 % CI 1.53–79.16]) and overall response rates (19/

26 versus 2/26, RR 9.5 [95 % CI 2.46–36.71]). An improve-

ment in patient self-assessed pruritus was documented in 21/26

patients treated with sulfasalazine and in only 2/26 patients

treated with placebo (RR 10.5 [95 % CI 2.74–40.29]). Only

Fig. 2 Oral glucocorticoids (GC) versus other therapies: complete response rates. CI confidence interval, df degrees of freedom, M–H Mantel–

Haenszel

L. Atzmony et al.

patients treated with sulfasalazine experienced adverse events

(8/26 versus 0/26, RR 17 [95 % CI 1.03–280.07]). Neverthe-

less, although the sulfasalazine group had a higher rate of

withdrawal due to adverse events, the difference in comparison

with the placebo group was not significant (3/26 versus 0/26,

RR 7 [95 % CI 0.38–280.07]).

3.2.5 Enoxaparin

One trial assessed the efficacy of subcutaneous enoxaparin

5 mg/week against oral prednisone 0.5 mg/kg/day for up to

8 weeks [14]. The complete response rate was significantly

lower for enoxaparin (8/27 versus 16/27, RR 0.95

[95 % CI 0.26–0.97]). Accordingly, the enoxaparin group

had a lesser reduction in itch (rated by a visual analog

scale) after treatment (2.84 versus 0.83, P = 0.005).

However, there were no significant between-group differ-

ences in the partial and overall response rates (10/27 versus

6/27, RR 1.67 [95 % CI 0.71–3.94] and 18/27 versus

22/27, RR 0.5 [95 % CI 0.13–1.91], respectively) or the

relapse rates (6/18 versus 9/22, RR 0.81 [95 % CI

0.36–1.86]). The number of patients with adverse events

was nonsignificantly lower in the enoxaparin group (1/27

versus 5/27, RR 0.2 [95 % CI 0.02–1.6]).

3.2.6 Methotrexate

One trial tested the efficacy and safety of oral methotrexate

10 mg/week against oral betamethasone for 12 weeks [9,

19]. As mentioned above, there was no between-group

difference in the complete response rate. The rates of some

adverse events were higher with betamethasone.

3.2.7 Metronidazole

A lower proportion of patients achieved overall ‘‘clinical

improvement’’ after 3 weeks of treatment with metron-

idazole than with PUVASOL and topical glucocorticoids

[21]. However, the planned intervention was not compa-

rable in terms of treatment duration. These results are

discussed in the section on phototherapy.

3.2.8 Hydroxychloroquine

A comparison of hydrochloroquine 400 mg/day with

griseofulvin 500 mg/day for 6 months yielded overall

response rates of 70 and 42.5 %, respectively [12]. These

results are discussed in the section on griseofulvin.

3.2.9 Dapsone

A nonrandomized case–control study comparing dapsone

150 mg/day with topical betamethasone 0.1 % reported a

nonsignificant between-group difference in the ‘‘good

response’’ rate after 3 months’ treatment (29/59 versus

10/25, RR 1.28 [95 % CI 0.75–2.12]) [22]. The authors

did not mention the type of betamethasone used. Of note,

most of the patients had classic lichen planus, but the

cohort also included patients with lichen actinicus, linear

lichen planus, lichen planus pigmentosus, and mucosal

lichen planus. Their responses were not analyzed sepa-

rately, so the possible effect of the different variants on

treatment in terms of direction and size could not be

determined.

3.3 Topical Therapy

Although topical glucocorticoids are used as a first-line

therapy in CLP, their efficacy has not been tested in a well-

designed RCT. Only three of the trials that were eligible for

the present review included topical glucocorticoids as a

treatment arm [16, 21, 22]. One assessed the efficacy of

calcipotriol 0.05 % against betamethasone valerate 0.1 %

ointment and found no difference between the groups in

‘‘lesion flattening’’ after 12 weeks (7/15 versus 8/16, RR 0.93

[95 % CI 0.45–1.94]) [16]. It is worth noting that the rate of

adverse events (irritation and increased pruritus severity) was

higher for calcipotriol, but the difference in comparison with

betamethasone dipropionate did not reach statistical signifi-

cance (3/15 versus 0/16, RR 7.43 [95 % CI 0.42–132.96]).

The second study of topical glucocorticoids reported a high

‘‘clinical improvement’’ rate for betamethasone dipropionate

0.05 % cream in comparison with oral metronidazole [21] at

12 weeks. However, this trial was subject to a high risk of

bias (see the sections on phototherapy and metronidazole).

The third study compared dapsone with topical betametha-

sone and showed a comparable ‘‘good response’’ (see section

on dapsone) [22].

One group examined the efficacy of KH 1060 ointment

(lexacalcitol)—a vitamin D3 analog—which had been

found to be approximately as effective as calcipotriol

ointment in the treatment of psoriasis [24], in comparison

with placebo [8]. After 8 weeks of treatment, there was no

between-group difference in the partial response rate (14/

49 versus 15/43, RR 0.95 [95 % CI 0.72–1.26]).

3.4 Strength of Body of Evidence

Table 3 details the results of the body-of-evidence analysis.

4 Discussion

In this systematic review, we aimed to evaluate the efficacy

of different treatment modalities for CLP. In contrast to

previous reviews, which included only small numbers of

Treatments for Cutaneous Lichen Planus

RCTs [2–4], our review included 16 trials, of which 11

were RCTs. The overall quality of evidence was very low

to moderate in terms of treatment efficacy. Acitretin, sul-

fasalazine, and griseofulvin were associated with higher

overall response rates than placebo. Acitretin and sul-

fasalazine caused significantly more adverse events than

placebo, but griseofulvin did not. NBUVB was more

effective than prednisolone in achieving a complete

Table 3 Quality of evidence for reviewed treatments

Intervention No. of

trials

Quality of

evidence

Effects

Oral prednisone 3 Moderate 10-day course did not significantly increase response and relapse rates versus placebo;

small sample size [10]

Medium dose for 8 weeks was more effective in achieving CR than enoxaparin [14]

Low dose for 6 weeks was less effective in achieving CR than NBUVB [13]

Topical GC 1 Moderate Flattening of lesions with betamethasone valerate 0.1 % ointment comparable to

calcipotriol ointment at 12 weeks (50 %); low PR rate; lichen planus type not

mentioned [16]

1 Low High and comparable ‘‘clinical improvement’’ rates with PUVASOL or betamethasone

dipropionate cream ? oral cetirizine for 12 weeks; permissive definition of

‘‘response’’ [21]

1 Very low Comparable ‘‘clinical improvement’’ rates with betamethasone 0.1 % (type not

mentioned) and dapsone [22]

Acitretin 30 mg/day for

8 weeks

1 Moderate More effective than placebo in achieving OR and reducing pruritus [11]

PO griseofulvin 500 mg/day

for 8 weeks

3 Moderate High response rate; more effective than placebo in achieving OR (not CR) [17, 18]

No relapses during 9 months’ follow-up [17]

In one low-quality RCT, response rate significantly lower than with hydroxychloroquine

[12]

Metronidazole 200 mg TID

for 3 weeks

1 Low Lower proportion of patients achieved overall ‘‘clinical improvement’’ after 3 weeks of

treatment with metronidazole than with PUVASOL or betamethasone dipropionate

0.1 % cream ? oral cetirizine [21]; short treatment duration compared with other

treatment arms

Sulfasalazine 2.5 mg/day for

6 weeks

1 Moderate More effective than placebo in achieving OR (not CR) and reducing pruritus [15]

Hydroxychloroquine

400 mg/day for 6 months

1 Low More effective than griseofulvin in achieving OR (not CR) [12]

SC enoxaparin 5 mg/week

for 8 weeks

1 Moderate Less effective than medium-dose prednisolone in achieving CR; comparable OR [14]

Dapsone 150 mg/day for

3 months

1 Very low ‘‘Clinical improvement’’ rates comparable to betamethasone 0.1 % (type not mentioned)

[22]

Methotrexate 10 mg/week

for 12 weeks

1 Low Nonsignificant difference in response rates and better safety profile compared with oral

betamethasone; quality downgraded to low because of inconclusiveness of results and

small sample size [9]

NBUVB 1 Moderate More effective than low-dose prednisone for 6 weeks in achieving CR [13]

1 Low OR rates comparable to PUVA; small sample size [23]

Topical PUVA 1 Low OR rate nonsignificantly higher with PUVA bath than with oral PUVA; small sample

size [20]

Oral PUVA 2 Low Effectiveness comparable to NBUVB; see above [23]

PUVASOL for 12 weeks 1 Low High and comparable ‘‘clinical improvement’’ rates with PUVASOL or betamethasone

dipropionate cream ? oral cetirizine for 12 weeks [21]

Topical vitamin D

derivatives

2 Moderate Comparable efficacy of calcipotriol 0.05 % and betamethasone valerate 0.1 % ointment

for 12 weeks [16]

Similar efficacy of KH 1060 and placebo for 8 weeks [8]

Quality assessment refers to OR/PR/CR/other reported response outcomes

CR complete response, GC glucocorticoid, NBUVB narrow-band ultraviolet B radiation, OR overall response, PO per os, PR partial response,

PUVA psoralen ? ultraviolet A photochemotherapy, PUVASOL psoralen ? sunlight exposure, RCT randomized controlled trial, SC subcuta-

neous, TID three times daily

L. Atzmony et al.

response, and prednisolone was more effective than

enoxaparin in achieving a complete response but not an

overall response. Hydroxychloroquine was more effective

than griseofulvin in achieving an overall response. Treat-

ment with low-dose methotrexate was effective, with a

good safety profile, but the difference in comparison with

oral betamethasone was not significant. Betamethasone

valerate 0.1 % ointment had comparable efficacy to cal-

cipotriol. Treatment with KH 1060 was not effective. In

nonrandomized controlled trials, oral PUVA had compa-

rable efficacy to a PUVA bath and NBUVB. PUVASOL

and betamethasone dipropionate 0.05 % cream were

effective in comparison with a short course of oral

metronidazole.

Because CLP may resolve spontaneously, the goals of

therapy are to shorten the time between onset and lesion

resolution and to reduce itching. Thus, the effectiveness of

therapy can be appraised by outcome measures of the

short-term response, time to response, and itch scores. In a

long-term follow-up study of patients with CLP, Irvine

et al. [25] found that the duration of the inflammatory

eruption was 1 year or less in 68 % of patients and

6 months or less in 42 %. In an older cohort, when potent

topical steroids were not available, 61 % of patients

showed lesion clearance within 12 months [26]. All but

one of the trials [12] included in this review reported the

short-term response. However, almost none reported the

time to response, and—with one exception [23]—those that

did used inappropriate statistical methods. In addition,

most of the trials did not use an itch score to evaluate

reduction of itch.

The decision on which CLP treatment to use for a

specific patient should be based on disease severity as well

as the patient’s medical background, the safety profile of

the treatment, and patient preferences. Although some

treatments were found to be less successful than others in

achieving a specific efficacy outcome, they may be con-

sidered first-line options in specific circumstances. For

example, subcutaneous enoxaparin was reported to be less

effective than oral prednisone in achieving a complete

response after 8 weeks of treatment. However, the overall

response rates in the two arms were comparable [14]. The

high overall response rate for enoxaparin was supported in

previous trials of lower methodological quality [27, 28].

Furthermore, its rate of adverse events was nonsignificantly

lower than that of prednisone [14]. The wide range of side

effects of systemic glucocorticoids is well known. There-

fore, a clinician may prefer enoxaparin for a patient with a

background of diabetes or osteoporosis.

Several limitations of this systematic review and the

current evidence should be noted. First, although there are

many variants of CLP with different clinical behaviors, this

review focused on classic (including generalized) CLP, and

the applicability of the results is restricted to these types.

Few patients with hypertrophic CLP and other variants

were included in a few trials [20–22], and we could not

analyze their responses to treatment separately. Given the

restricted number of published comparative trials, we

compromised our analysis by including those with unclear

definitions of ‘‘response’’. As this reporting bias can impact

the outcome effect size and even direction, in such cases,

we downgraded the quality of evidence. Second, we pre-

sented primarily analyses for modified-ITT and clinically

evaluable populations. In noninferiority studies, the ITT

results might bias the overall results toward no difference

[29]. Nevertheless, we opted for ITT analysis because in

clinical practice, we are interested in the results for all

patients treated (the ITT population), and derivation of the

per-protocol population in the included trials could have

led to other biases. Third, since most of the included trials

had a short follow-up and a small sample size, we could not

appropriately assess adverse drug effects. The small sample

size also led to inconclusive results in some noninferiority

trials [9, 14].

5 Conclusions

Taking the limited quality of the evidence into consider-

ation, several treatment modalities were found to be

effective for CLP—specifically, for classic or generalized

CLP. Systemic treatments with acitretin, sulfasalazine,

and griseofulvin were associated with increased overall

response rates compared with placebo. Hydroxychloro-

quine was more effective than griseofulvin in achieving

an overall response. NBUVB was more effective than

6 weeks’ low-dose prednisolone in achieving a complete

response, and prednisolone was more effective than

enoxaparin. Methotrexate was effective, with a non-

significant difference in the complete response rate in

comparison with oral betamethasone. Topical treatment

with betamethasone valerate 0.1 % ointment had compa-

rable efficacy to calcipotriol ointment. Oral PUVA had

comparable efficacy to a PUVA bath and NBUVB.

PUVASOL and betamethasone dipropionate 0.05 %

cream were effective relative to a short course of oral

metronidazole. Further large-scale RCTs are warranted to

investigate the efficacy of topical glucocorticoids—the

current first-line therapy—as well as other treatment

modalities. Future studies should focus on both classic

CLP and different variants of CLP, with separate end-

points for each variant. Uniform definitions of ‘‘response’’

should be used, including evaluation of pruritus and time

to response.

Treatments for Cutaneous Lichen Planus

Compliance with Ethical Standards

Conflict of interest The authors—Lihi Atzmony, Ofer Reiter,

Emmilia Hodak, Michael Gdalevich, and Daniel Mimouni—have no

conflicts of interest that are directly relevant to this work.

Funding No funding was received for this work.

References

1. Shiohara T, Kano Y. Lichen planus and lichenoid dermatoses. In:

Bolognia JL, Jorizzo JL, Schaffer JV, editors. Dermatology. 3rd

ed. London: Mosby Elsevier; 2012. p. 183–96.

2. Cribier B, Frances C, Chosidow O. Treatment of lichen planus:

an evidence-based medicine analysis of efficacy. Arch Dermatol.

1998;134:1521–30.

3. Antiga E, Caproni M, Parodi A, Cianchini G, Fabbri P. Treatment

of cutaneous lichen planus: an evidence based analysis of efficacy

by the Italian Group for Cutaneous Immunopathology. G Ital

Dermatol Venereol. 2014;149:719–26.

4. Fazel N. Cutaneous lichen planus: a systematic review of treat-

ments. J Dermatolog Treat. 2014;9:1–4.

5. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group.

Preferred reporting items for systematic reviews and meta-anal-

yses: the PRISMA statement. PLoS Med. 2009;6:e1000097.

6. Downs SH, Black N. The feasibility of creating a checklist for the

assessment of the methodological quality both of randomised and

non-randomised studies of health care interventions. J Epidemiol

Community Health. 1998;52:377–84.

7. Schunemann H, Hill S, Guyatt G, Akl EA, Ahmed F. The

GRADE approach and Bradford Hill’s criteria for causation.

J Epidemiol Community Health. 2011;65:392–5.

8. Bouloc A, Revuz J, Bagot M, Wechsler J, Natta P. KH 1060 for

the treatment of lichen planus: a multicenter, randomized, dou-

ble-blind, vehicle-controlled study. Arch Dermatol.

2000;136:1272.

9. Hazra SC, Choudhury AM, Khondker L, Khan SI. Comparative

efficacy of methotrexate and mini pulse betamethasone in the

treatment of lichen planus. Mymensingh Med J. 2013;22:787–97.

10. Kellett JK, Ead RD. Treatment of lichen planus with a short

course of oral prednisolone. Br J Dermatol. 1990;123:550–1.

11. Laurberg G, Geiger JM, Hjorth N, Holm P, Hou-Jensen K,

Jacobsen KU, et al. Treatment of lichen planus with acitretin: a

double-blind, placebo-controlled study in 65 patients. J Am Acad

Dermatol. 1991;24:434–7.

12. Bhuiyan I, Wahab M, Ali A, Sultana A, Siddique R, Hawlader A,

et al. Comparative efficacy of hydroxychloroquine and griseo-

fulvin in the treatment of lichen planus. J Pakistan Assoc Der-

matol. 2010;20:79–83.

13. Iraji F, Faghihi G, Asilian A, Siadat AH, Larijani FT, Akbari M.

Comparison of the narrow band UVB versus systemic corticos-

teroids in the treatment of lichen planus: a randomized clinical

trial. J Res Med Sci. 2011;16:1578–82.

14. Iraji F, Asilian A, Saeidi A, Siadat AH, Saeidi AR, Hassanzadeh

A. Comparison of therapeutic effect of low-dose low-molecular-

weight heparin (enoxaparin) vs. oral prednisone in treatment of

patients with lichen planus: a clinical trial. Adv. Biomed Res.

2013;2:76.

15. Omidian M, Ayoobi A, Mapar MA, Feily A, Cheraghian B.

Efficacy of sulfasalazine in the treatment of generalized lichen

planus: randomized double-blinded clinical trial on 52 patients.

J Eur Acad Dermatol Venereol. 2010;24:1051–4.

16. Theng CT, Tan SH, Goh CL, Suresh S, Wong HB, Machin D,

Singapore Lichen Planus Study Group. A randomized controlled

trial to compare calcipotriol with betamethasone valerate for the

treatment of cutaneous lichen planus. J Dermatolog Treat.

2004;15:141–5.

17. Sehgal VN, Abraham GJ, Malik GB. Griseofulvin therapy in

lichen planus: a double-blind controlled trial. Br J Dermatol.

1972;87:383–5.

18. Sehgal VN, Bikhchandani R, Koranne RV, Nayar M, Saxena

HM. Histopathological evaluation of griseofulvin therapy in

lichen planus: a double-blind controlled study. Dermatologica.

1980;161:22–7.

19. Hazra SC, Choudhury AM, Asaduzzaman AT, Paul HK. Adverse

outcome of methotrexate and mini pulse betamethasone in the

treatment of lichen planus. Bangladesh Med Res Counc Bull.

2013;39:22–7.

20. Helander I, Jansen CT, Meurman L. Long-term efficacy of PUVA

treatment in lichen planus: comparison of oral and external

methoxsalen regimens. Photodermatol. 1987;4:265–8.

21. Sharma L, Mishra MK. A comparative study of PUVASOL

therapy in lichen planus. Indian J Dermatol Venereol Leprol.

2003;69:212–3.

22. Chopra A, Mittal RR, Kaur B. Dapsone versus corticosteroids in

lichen planus. Indian J Dermatol Venereol Leprol. 1999;65:66–8.

23. Wackernagel A, Legat FJ, Hofer A, Quehenberger F, Kerl H,

Wolf P. Psoralen plus UVA vs. UVB-311 nm for the treatment of

lichen planus. Photodermatol Photoimmunol Photomed.

2007;23:15–9.

24. Kragballe K, Dam TN, Hansen ER, Baadsgaard O, Grønhøj

Larsen F, Søndergaard J, et al. Efficacy and safety of the 20-epi-

vitamin D3 analogue KH 1060 in the topical therapy of psoriasis:

results of a dose-ranging study. Acta Derm Venereol.

1994;74(5):398–402.

25. Irvine C, Irvine F, Champion RH. Long-term follow-up of lichen

planus. Acta Derm Venereol. 1991;71:242–4.

26. Samman PD. Lichen planus: an analysis of 200 cases. Trans St

Johns Hosp Dermatol Soc. 1961;46:36–8.

27. Hodak E, Yosipovitch G, David M, Ingber A, Chorev L, Lider O,

et al. Low-dose low-molecular-weight heparin (enoxaparin) is

beneficial in lichen planus: a preliminary report. J Am Acad

Dermatol. 1998;38:564–8.

28. Khan SI, Khondker L. Safety of low molecular weight heparin in

the treatment of cutaneous lichen planus. Mymensingh Med J.

2012;21:668–73.

29. Piaggio G, Elbourne DR, Altman DG, Pocock SJ, Evans SJ,

CONSORT Group. Reporting of noninferiority and equivalence

randomized trials: an extension of the CONSORT statement.

JAMA. 2006;295:1152–60.

L. Atzmony et al.