Interventions for treating oral lichen planus: corticosteroid therapies

Giovanni Lodi; Maddalena Manfredi; Valeria Mercadante; Ruth Murphy; Marco Carrozzo

doi:10.1002/14651858.CD001168.pub3

. 2020 Feb 28;2020(2):CD001168. doi: 10.1002/14651858.CD001168.pub3

Interventions for treating oral lichen planus: corticosteroid therapies

Giovanni Lodi ¹, Maddalena Manfredi ², Valeria Mercadante ³, Ruth Murphy ⁴, Marco Carrozzo ^5,^✉

Editor: Cochrane Oral Health Group

PMCID: PMC7047223 PMID: 32108333

Abstract

Background

Oral lichen planus (OLP) is a relatively common chronic T cell‐mediated disease, which can cause significant pain, particularly in its erosive or ulcerative forms. As pain is the indication for treatment of OLP, pain resolution is the primary outcome for this review. This review is an update of a version last published in 2011, but focuses on the evidence for corticosteroid treatment only. A second review considering non‐corticosteroid treatments is in progress.

Objectives

To assess the effects and safety of corticosteroids, in any formulation, for treating people with symptoms of oral lichen planus.

Search methods

Cochrane Oral Health's Information Specialist searched the following databases to 25 February 2019: Cochrane Oral Health's Trials Register, CENTRAL (2019, Issue 1), MEDLINE Ovid, and Embase Ovid. ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform were searched for ongoing trials. There were no restrictions on language or date of publication.

Selection criteria

We considered randomised controlled clinical trials (RCTs) of any local or systemic corticosteroid treatment compared with a placebo, a calcineurin inhibitor, another corticosteroid, any other local or systemic (or both) drug, or the same corticosteroid plus an adjunctive treatment.

Data collection and analysis

Three review authors independently scanned the titles and abstracts of all reports identified, and assessed risk of bias using the Cochrane tool and extracted data from included studies. For dichotomous outcomes, we expressed the estimates of effects of an intervention as risk ratios (RR), with 95% confidence intervals (CI). For continuous outcomes, we used mean differences (MD) and 95% CI. The statistical unit of analysis was the participant. We conducted meta‐analyses only with studies of similar comparisons reporting the same outcome measures. We assessed the overall certainty of the evidence using GRADE.

Main results

We included 35 studies (1474 participants) in this review. We assessed seven studies at low risk of bias overall, 11 at unclear and the remaining 17 studies at high risk of bias. We present results for our main outcomes, pain and clinical resolution measured at the end of the treatment course (between one week and six months), and adverse effects. The limited evidence available for comparisons between different corticosteroids, and corticosteroids versus alternative or adjunctive treatments is presented in the full review.

Corticosteroids versus placebo

Three studies evaluated the effectiveness and safety of topical corticosteroids in an adhesive base compared to placebo. We were able to combine two studies in meta‐analyses, one evaluating clobetasol propionate and the other flucinonide. We found low‐certainty evidence that pain may be more likely to be resolved when using a topical corticosteroid rather than a placebo (RR 1.91, 95% CI 1.08 to 3.36; 2 studies, 72 participants; I² = 0%). The results for clinical effect of treatment and adverse effects were inconclusive (clinical resolution: RR 6.00, 95% CI 0.76 to 47.58; 2 studies, 72 participants; I² = 0%; very low‐certainty evidence; adverse effects RR 1.48, 95% 0.48 to 4.56; 3 studies, 88 participants, I² = 0%, very low‐certainty evidence).

Corticosteroids versus calcineurin inhibitors

Three studies compared topical clobetasol propionate versus topical tacrolimus. We found very low‐certainty evidence regarding any difference between tacrolimus and clobetasol for the outcomes pain resolution (RR 0.45, 95% CI 0.24 to 0.88; 2 studies, 100 participants; I² = 80%), clinical resolution (RR 0.61, 95% CI 0.38 to 0.99; 2 studies, 52 participants; I² = 95%) and adverse effects (RR 0.05, 95% CI 0.00 to 0.83; 2 studies, 100 participants; very low‐certainty evidence) .

One study (39 participants) compared topical clobetasol and ciclosporin, and provided only very low‐certainty evidence regarding the rate of clinical resolution with clobetasol (RR 3.16, 95% CI 1.00 to 9.93), pain resolution (RR 2.11, 95% CI 0.76 to 5.86) and adverse effects (RR 6.32, 95% CI 0.84 to 47.69).

Two studies (60 participants) that compared triamcinolone and tacrolimus found uncertain evidence regarding the rate of clinical resolution (RR 0.86, 95% CI 0.55 to 1.35; very low‐certainty evidence) and that there may be a lower rate of adverse effects in the triamcinolone group (RR 0.47, 95% CI 0.22 to 0.99; low‐certainty evidence). These studies did not report on pain resolution.

Authors' conclusions

Corticosteroids have been first line for the treatment of OLP. This review found that these drugs, delivered topically as adhesive gels or similar preparations, may be more effective than placebo for reducing the pain of symptomatic OLP; however, with the small number of studies and participants, our confidence in the reliability of this finding is low. The results for clinical response were inconclusive, and we are uncertain about adverse effects. Very low‐certainty evidence suggests that calcineurin inhibitors, specifically tacrolimus, may be more effective at resolving pain than corticosteroids, although there is some uncertainty about adverse effects and clinical response to tacrolimus showed conflicting results.

Plain language summary

Corticosteroids for treating oral lichen planus

Review question

Are corticosteroids effective and safe for the treatment of oral lichen planus that is causing pain?

Background

Oral lichen planus is a common condition that can cause long‐term, painful areas on the lining of the mouth. Usual treatment is with drugs known as corticosteroids applied directly to the painful areas (topically), or taken internally (systemically). Treatment aims to reduce pain and improve healing of the mouth, but there is no cure for the disease.

Study characteristics

The evidence in this review is up‐to‐date as of 25 February 2019. We included 35 randomised controlled trials (clinical studies where people are randomly put into one of two or more treatment groups) with 1474 participants, which tested several different corticosteroids, mostly delivered topically (on the skin). Corticosteroids were compared with one of the following: a placebo (a treatment that resembled the corticosteroid but had no active ingredient); a medicine from a category called calcineurin inhibitors; another medicine type; another corticosteroid or mode of delivery; the same corticosteroid plus an extra treatment; or an alternative treatment. Treatments were given for between one week and six months, with side effects measured throughout the treatment, and pain and healing measured at the end of treatment.

Key results

Results from two studies showed that topical corticosteroids (e.g. clobetasol propionate, flucinonide, betamethasone and triamcinolone acetonide), when applied to the mouth in a sticky cream, may be effective in reducing and stopping pain. We do not have the evidence that topical corticosteroids can eliminate the oral lichen planus lesions, and we are uncertain about side effects.

We found no consistent evidence that any particular corticosteroid was better than any other.

Three studies using another topical medicine called tacrolimus (a calcineurin inhibitor) found that this medicine may be more effective than corticosteroids, but may be more likely to cause mild side effects.

Available evidence comparing corticosteroids with other treatments is very limited.

Reliability of the evidence

The reliability of most of the evidence is very low, so we cannot be sure about the findings and future research may lead us to different conclusions.

Conclusion

The available evidence suggests that topical corticosteroids may be effective for treating painful oral lichen planus, but our confidence in these findings is limited as there were only a small number of studies and participants. There is some evidence that tacrolimus may be more effective than a corticosteroid, but evidence on negative side effects is inconclusive.

Summary of findings

Background

Description of the condition

Oral lichen planus (OLP) is a chronic disorder of the oral cavity that rarely undergoes spontaneous remission. Despite the lack of good epidemiological data, OLP is thought to be relatively common, affecting approximately 1% to 2% of the population, mainly middle‐aged adults (Alrashdan 2016). Women are slightly more likely than men to have this condition. The most commonly affected sites are the buccal mucosa bilaterally, the borders and dorsum of the tongue, and the gingiva. The palate (either hard or soft), the lips and the floor of the mouth are rarely involved.

Typical OLP clinical features are represented by bilaterally located white papules that enlarge and coalesce to form reticulations, the so‐called Wickham's striae (Carrozzo 1999), which are rarely symptomatic. In contrast, erythematous and erosive or ulcerative lesions can cause varying degrees of discomfort. Symptomatic OLP is relatively frequent and can significantly impair quality of life (López‐Jornet 2010; Tadakamadla 2015).

Moreover, the disease has a fluctuating course with apparent spontaneous exacerbations and improvements in disease activity within an individual patient.

Current evidence suggests that people with OLP have an increased risk of developing oral squamous cell carcinoma (SCC) (Eisen 2002; Gonzalez‐Moles 2008), and this has to be considered when planning therapeutic interventions (Aghbari 2017). However, this topic will not be addressed in this review.

Clinical appearance alone, particularly when showing the 'classic' reticular form, may sometimes allow a definitive diagnosis. However, given the chronic course of the disease, the sometimes pleomorphic clinical manifestations, and the common long‐term treatment and monitoring of people with OLP, biopsy is a prudent – yet still controversial – clinical practice. Inappropriate diagnosis is a notable cause of therapy failure, so histopathological confirmation of OLP is helpful before starting an active treatment. Histopathology can be subjective and non‐specific (Van der Meij 2003), but it can be useful to exclude dysplasia and SCC. When exclusive gingival or predominantly erosive or ulcerative lesions are present, immunological tests are warranted to achieve a proper diagnosis.

OLP is probably a T cell‐mediated immunological reaction to an induced antigenic change in the oral mucosa in predisposed people. An early event in OLP is the genetically driven enhanced production of Th1 cytokines, particularly tumour necrosis factor‐alpha (TNF‐α) and interferon‐gamma (IFN‐γ) (Carrozzo 2004). Studies of T cell receptor variable region genes have highlighted that OLP is likely to be the common outcome of a limited combination of extrinsic antigens, altered self‐antigens or super antigens (Thomas 1997). In a minority of people, aetiological factors can be identified and they are usually drugs, dental materials and infectious agents, especially hepatitis C virus infection (Lodi 2005a; Lodi 2010).

Description of the intervention

Various treatment regimens have been employed to treat ulcerative lesions, and, more importantly, to reduce the associated pain, though a definitive cure for OLP has not yet been achieved (Lodi 2005b). The primary goal of treatment of symptomatic OLP is the reduction and preferably elimination of pain associated with the lesions.

How the intervention might work

Because OLP is considered a T cell‐mediated disease associated with a Th1 imbalance of cytokine production, most of the therapeutic interventions have aimed to target the inflammatory pathway underlying OLP. In particular, local suppression of T cells and a reduction in the release of cytokines such as TNF‐α and IFN‐γ are highly regarded in OLP management. As a result, the mainstay medications in OLP management are anti‐inflammatory drugs. The most commonly used anti‐inflammatory medication is glucocorticosteroids, commonly called corticosteroids. Around 1950, topical glucocorticosteroids (TGCs) were employed to treat skin inflammatory disease; the use of TGCs for mouth diseases, including OLP, started around a decade later (Zegarelli 1960). TGCs have a multiplicity of actions: anti‐inflammatory, immunomodulatory, vasoconstrictor, and they can inhibit the activity of several cytokines following inactivation of specific transcription factors such as activator protein 1 (AP‐1) and nuclear factor kappa B (NFκB) (Ahluwalia 1998). Specifically, the analgesic effect of corticosteroids is likely related to their effect on the inflammatory pathway underlying OLP and its beneficial effect on mucosal healing and integrity.

Why it is important to do this review

Symptomatic OLP is a relatively common, painful oral disorder that can significantly impair quality of life (Tadakamadla 2015). Because of its chronic nature and lack of an apparent cause, a definitive cure is very difficult to achieve. Current treatments aim to reduce pain and to heal erosive and ulcerative lesions. Most published reviews on the topic suggest the use of topical drugs, mainly TGCs (Al‐Hashimi 2007; Carrozzo 1999; Carrozzo 2009; Cribier 1998; Eisen 2005; Lodi 2005b); however, the previous version of this Cochrane Review provided only weak evidence for the superiority of any interventions over placebo for palliation of symptomatic OLP (Other published versions of this review). As we were aware of an increasing number of randomised controlled trials (RCTs) evaluating corticosteroids, we updated the review, focusing on these interventions. A further review on non‐corticosteroids is also being produced.

Objectives

To assess the effects and safety of corticosteroids, in any formulation, for treating people with symptoms of oral lichen planus.

Methods

Criteria for considering studies for this review

Types of studies

We included RCTs. We excluded quasi‐randomised trials (where treatment assignment was by alternating sequence, date of birth, registration number or some other such non‐random method).

Types of participants

We included participants satisfying the following criteria.

Having a clinical and histological diagnosis of OLP.
Having painful symptoms associated with OLP.
Not concurrently receiving any other treatment for OLP or treatment likely to modify their OLP (e.g. systemic steroids, antifungals or immunosuppressants).

For people being treated for both skin and OLP, we extracted only the OLP data; if this was not possible, we excluded the study.

Types of interventions

Any local or systemic corticosteroid treatment compared with a placebo, a calcineurin inhibitor, another corticosteroid, any other local or systemic treatment, or the same corticosteroid plus an adjunctive treatment.

Types of outcome measures

Primary outcomes

Pain (score and resolution) as assessed by participants (measured at the end of the treatment course).

Secondary outcomes

Clinical response (score and resolution of the disease) in terms of changes in the extension and severity (degree of erosion, erythema and reticulation) as assessed by clinicians (measured at the end of the treatment course).
Adverse effects, including clinical candidiasis or other toxic and side effects (measured at any time point).

Search methods for identification of studies

Electronic searches

Cochrane Oral Health's Information Specialist conducted systematic searches in the following databases. The search was inclusive of RCTs and controlled clinical trials, but the latter were filtered out during the selection process. There were no language, publication year or publication status restrictions.

Cochrane Oral Health's Trials Register (searched 25 February 2019; Appendix 1).
Cochrane Central Register of Controlled Trials (CENTRAL; 2019, Issue 1) in the Cochrane Library (searched 25 February 2019; Appendix 2).
MEDLINE Ovid (1946 to 25 February 2019; Appendix 3).
Embase Ovid (1980 to 25 February 2019; Appendix 4).

Subject strategies were modelled on the search strategy designed for MEDLINE Ovid. Where appropriate, they were combined with subject strategy adaptations of the highly sensitive search strategy designed by Cochrane for identifying RCTs and controlled clinical trials as described in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions (Lefebvre 2011).

Searching other resources

The following trial registries were searched for ongoing studies:

US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (clinicaltrials.gov; searched 25 February 2019; Appendix 5);
World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch; searched 25 February 2019; Appendix 6).

We also checked the reference lists of identified publications for relevant studies, and contacted authors to identify missing and unreported trials.

We checked that none of the included studies in this review were retracted due to error or fraud.

We did not perform a separate search for adverse effects of interventions used; we considered adverse effects described in included studies only.

Data collection and analysis

Selection of studies

Three review authors independently scanned titles and abstracts (when available) of all reports identified. The search was designed to be sensitive and include controlled clinical trials, these were filtered out early in the selection process if they were not randomised. For studies appearing to meet the inclusion criteria, or when there was insufficient information in the title and abstract to make a clear decision, we obtained the full reports and all review authors independently assessed them to establish if they met inclusion criteria. We resolved disagreements by discussion. We recorded studies that we rejected at this or subsequent stages in the Characteristics of excluded studies table, along with the reasons for exclusion.

Data extraction and management

At least two review authors extracted data from all studies meeting the inclusion criteria, using a specially designed form. We recorded the characteristics of the trial participants, interventions and outcomes in the Characteristics of included studies table. The studies measured the effects of treatment on pain and clinical presentation using scales and scoring systems, which were often significantly different and difficult to compare. In order to increase the amount of comparable data, we decided to record the number of participants who did not receive benefit in terms of symptoms (pain) and clinical signs.

Assessment of risk of bias in included studies

All review authors independently assessed the risk of bias of the included trials. All review authors independently assessed the full‐text papers, unblinded, and resolved disagreements through discussion and consensus. We used the recommended tool for assessing risk of bias in studies included in Cochrane Reviews (Higgins 2011). It is a two‐part tool, addressing seven specific domains as follows:

random sequence generation (selection bias);
allocation concealment (selection bias);
blinding of participants and personnel (performance bias);
blinding of outcome assessment (detection bias);
incomplete outcome data (attrition bias);
selective reporting (reporting bias);
other bias.

Each domain in the tool includes one or more specific entries in a 'Risk of bias' table. Within each entry, the first part of the tool describes what was reported to have happened in the study, in sufficient detail to support a judgement about the risk of bias. The second part of the tool assigns a judgement relating to the risk of bias for that entry. This is achieved by assigning a judgement of 'low', 'high' or 'unclear' risk of bias.

After taking into account the additional information provided by trial authors, we categorised trials as:

overall low risk of bias if low risk of bias for all key domains;
overall unclear risk of bias if unclear risk of bias for one or more key domains; or
overall high risk of bias if high risk of bias for one or more key domains.

We completed a 'Risk of bias' table for each included study (see Characteristics of included studies table), and presented the results graphically by study and by domain across all studies.

Measures of treatment effect

For dichotomous outcomes, we expressed the estimates of effects of an intervention as risk ratios (RR) or odds ratios (OR) if paired, together with 95% confidence intervals (CIs). For continuous outcomes, we used mean differences (MD) and standard deviation (SD) for each group in order to express the estimate of effect as MD with 95% CI. If studies reported continuous outcomes on different scales, we planned to use standardised mean difference (SMD) to pool these data in meta‐analyses. For paired data (split‐mouth studies), we used the generic inverse variance method (Higgins 2011).

Unit of analysis issues

The statistical unit of analysis was the participant.

For studies with more than two control arms, we selected the one we considered most appropriate to compare.

We intended to analyse split‐mouth and cross‐over trials. Where the intraclass correlation was not provided for cross‐over and split‐mouth trials, we estimated this as 0.5.

Dealing with missing data

We contacted trialists to ask them to supply missing information and to clarify points.

Assessment of heterogeneity

We assessed the significance of any discrepancies in the estimates of the treatment effects from the different trials by means of Cochrane's test for heterogeneity and the I² statistic, which describes the percentage total variation across studies that is due to heterogeneity rather than chance. We considered heterogeneity to be statistically significant if the P value was less than 0.1. A rough guide to the interpretation of the I² statistic given in the Cochrane Handbook for Systematic Reviews of Interventions is: 0% to 40% might not be important, 30% to 60% may represent moderate heterogeneity, 50% to 90% may represent substantial heterogeneity and 75% to 100% is considerable heterogeneity (Higgins 2011).

Assessment of reporting biases

We attempted to minimise reporting biases by conducting a thorough search of multiple sources including trial registries, and efforts to identify unpublished trials and non‐English language publications.

Data synthesis

Where studies comparing similar interventions reported the same outcome measures, we combined the data in meta‐analyses. We combined RRs for dichotomous data, and MDs for continuous data, using fixed‐effect models unless there were more than three studies being combined. We dichotomised four‐ or five‐step rating scales as appropriate. If studies reported continuous outcomes on different scales, we planned to use SMD to pool these data in meta‐analyses.

Subgroup analysis and investigation of heterogeneity

If there were sufficient studies, we planned to assess clinical heterogeneity by examining the characteristics of participants included in the studies.

Sensitivity analysis

If there were sufficient studies, we planned to undertake sensitivity analyses to examine the effect of the study risk of bias assessment on the overall estimates of effect.

'Summary of findings' table

We created 'Summary of findings' tables for the comparisons of treatments considered 'first choice' and routinely adopted in clinical practice, and presented summary information for primary outcomes, in particular pain resolution, clinical resolution and adverse effects. At least two review authors (of GL, MM, MC, VM) independently assessed the certainty of evidence using GRADE criteria (Schünemann 2017), which considers a body of RCT evidence to provide high‐certainty evidence unless 'downgraded' by one, two or three levels (to moderate, low or very low certainty, respectively) on the basis of problems in study design, imprecision, inconsistency, indirectness or publication bias.

Results

Description of studies

See Characteristics of included studies table.

Results of the search

We identified 1030 records and rejected 943 on the basis of title or abstract. We considered 87 articles in full text and excluded 36 of these (we also excluded 11 studies that had been included in the previous version of the review as these will be part of our sister review on non‐corticosteroids). Two studies are awaiting classification and we found 12 ongoing studies. Therefore, we included 37 articles (35 RCTs) in this review. See Figure 1.

Study flow diagram of searches for this update.

Included studies

For a summary of the characteristics of each of the included studies, see the Characteristics of included studies tables.

Characteristics of the trials

Design

Twenty‐nine trials used a two‐arm parallel design; four used a three‐arm parallel design (Hesen 2017; Siponen 2017; Sivaraman 2016; Thomas 2017), though we used only two arms from Hesen 2017 and Thomas 2017; one used a two‐arm cross‐over design (Hegarty 2002); and one used a split‐mouth design (Amanat 2014).

The total number of participants included in the trials was 1474, with the number per study ranging from 20 to 139.

Setting

Seven studies were conducted in Iran (Amanat 2014; Amirchaghmaghi 2016; Bakhtiari 2017; Ghabanchi 2009; Gorouhi 2007; Kia 2015; Pakfetrat 2015), six in Italy (Arduino 2018; Campisi 2004; Carbone 2009; Conrotto 2006; Corrocher 2008; Lodi 2007), five in China (Fu 2012; Liu 2013; Wei 2003; Xiong 2009; Xu 2002), four in India (Arunkumar 2015; Malhotra 2008; Sivaraman 2016; Thomas 2017), four in Egypt (Ezzatt 2019; Hashem 2019; Hesen 2017; Mostafa 2017), two in the Netherlands (Laeijendecker 2006; Voute 1993), one in the US (Chainani‐Wu 2007), one in Sweden (Rodstrom 1994), one in the UK (Hegarty 2002), one in Brazil (Dillenburg 2014), one in Sri Lanka (Hettiarachchi 2017), one in Finland (Siponen 2017), and one was an Asian multicentre study (Singapore, South Korea, India, Thailand) (Yoke 2006). All the studies were conducted in university clinics or hospitals.

Funding

Companies provided drugs to four trials (Conrotto 2006; Gorouhi 2007; Voute 1993; Yoke 2006); one trial received support for the multicentric co‐ordination and study drug from the sponsor (Yoke 2006); two received support from the principal investigator (Arduino 2018; Ezzatt 2019); and 10 received support from institutional funding bodies (Amanat 2014; Amirchaghmaghi 2016; Chainani‐Wu 2007; Dillenburg 2014; Ghabanchi 2009; Hettiarachchi 2017; Lodi 2007; Pakfetrat 2015; Siponen 2017; Xiong 2009).

Characteristics of the interventions

See Table 3.

1. Characteristics of the interventions.

Comparison	Delivery method	Interventions	Study
Corticosteroids vs placebo	Topical	Fluocinonide 0.025% in 40% hypromellose ointment in white soft paraffin vs placebo	Voute 1993
	Topical	Triamcinolone acetonide ointment 0.1% vs placebo	Siponen 2017
	Topical	Clobetasol propionate 0.025% in 4% hydroxyethyl cellulose gel vs placebo^a	Arduino 2018
Corticosteroids vs calcineurin inhibitors	Topical	Triamcinolone acetonide 0.1% paste vs pimecrolimus 1% paste	Arunkumar 2015
	Topical	Clobetasol propionate 0.025% in 4% hydroxyethyl cellulose gel vs ciclosporin 1.5% in 4% hydroxyethyl cellulose gel^a	Conrotto 2006
	Topical	Clobetasol propionate 0.05% ointment vs tacrolimus 0.1% ointment	Corrocher 2008
	Topical	Betamethasone 0.1% gel vs pimecrolimus 1% gel	Ezzatt 2019
	Topical	Clobetasol propionate 0.05% cream vs tacrolimus 0.1% cream	Hettiarachchi 2017
	Topical	Clobetasol propionate 0.05% vs tacrolimus 0.03%	Sivaraman 2016
	Topical	Triamcinolone 0.1% paste vs pimecrolimus 1% cream	Gorouhi 2007
	Topical	Triamcinolone acetonide 0.1% in Orabase vs pimecrolimus 1% cream	Pakfetrat 2015
	Topical	Triamcinolone acetonide 0.1% in hypromellose 20% ointment vs tacrolimus 1% ointment	Laeijendecker 2006
	Topical	Triamcinolone acetonide ointment 0.1% vs tacrolimus 0.1% ointment	Siponen 2017
	Topical	Triamcinolone acetonide 0.1% vs tacrolimus 0.03%	Sivaraman 2016
	Topical	Triamcinolone acetonide 0.1% in Orabase vs ciclosporin solution 0.1%	Yoke 2006
Corticosteroid A vs corticosteroid B	Topical	Clobetasol propionate in microspheres 0.025% vs clobetasol propionate 0.025% in a dispersion of a lipophilic ointment in a hydrophilic phase	Campisi 2004
	Topical	Clobetasol priopionate 0.05% in 4% hydroxyethyl cellulose vs clobetasol priopionate 0.025% in 4% hydroxyethyl cellulose^a	Carbone 2009
	Topical	Clobetasol propionate 0.05% ointment in Orabase vs triamcinolone acetonide 0.1% ointment in Orabase	Rodstrom 1994
	Topical	Clobetasol propionate 0.05% vs triamcinolone acetonide 0.1%	Sivaraman 2016
	Injected locally	Betamethasone dipropionate 5 mg + betamethasone disodium phosphate 2 mg/mL intralesional injection vs triamcinolone acetonide 8 mg, intralesional injection	Liu 2013
	Systemic	Betamethasone 5 mg daily orally vs triamcinolone acetonide 0.1% paste	Malhotra 2008
	Topical	Prednisolone 5 mg mucoadhesive tablet vs triamcinolone acetonide 0.1% paste	Ghabanchi 2009
	Topical	Fluticasone spray 50 μg vs betamethasone sodium phosphate 500 μg oral rinses	Hegarty 2002
Corticosteroids vs other treatments	Topical	Triamcinolone acetonide paste 0.1% vs curcumin paste 5%	Kia 2015
	Topical	Triamcinolone acetonide paste 0.1% vs Curenext Oral Gel containing curcuma longa extracts 10 mg	Thomas 2017
	Topical	Dexamethasone mouthwash 0.5 mg oral rinses vs photodynamic therapy mediated by methylene blue	Bakhtiari 2017
	Topical	Triamcinolone acetonide 0.1% paste vs hyaluronic acid preparation 0.2% paste	Hashem 2019
	Topical	Triamcinolone acetonide 0.1% in Orabase vs photodynamic therapy mediated by methylene blue	Mostafa 2017
	Topical	Triamcinolone acetonide ointment 0.1% in Orabase vs cryotherapy	Amanat 2014
	Topical	Clobetasol propionate gel 0.05% vs laser diode^b	Dillenburg 2014
	Injected locally	Triamcinolone acetonide 10 mg intralesional injection vs Bacillus Calmette‐Guerin polysaccharide nucleic acid, 0.5 mL intralesional injection	Xiong 2009
	Topical	Dexamethasone paste 0.043% vs amlexanox paste 250 mg	Fu 2012
Adjunctive treatment to corticosteroids	Topical	Clobetasol propionate gel 0.05% + miconazole 2% gel vs clobetasol propionate gel 0.05% + placebo gel	Lodi 2007
	Topical	Dexamethasone paste and mycostatin paste alternatively vs dexamethasone paste	Wei 2003
	Topical	Dexamethasone mouthwash 0.5 mg oral rinses + curcumin 250 mg tablets vs dexamethasone mouthwash 0.5 mg oral rinses + placebo^b	Amirchaghmaghi 2016
	Topical	Triamcinolone acetonide 0.1% oral paste + glucosamine sulphate 1500 mg vs triamcinolone acetonide 0.1% oral paste	Hesen 2017
	Systemic	Prednisone 60 mg + curcuminoids 2000 mg daily vs prednisone 60 mg + placebo	Chainani‐Wu 2007
	Systemic	Herbal topical and decoction + prednisone 5–10 mg 3 times daily chlorphenamine 4 mg 3 times daily, vitamin C 0.1 g 3 times daily vs prednisone 5–10 mg 3 times daily, chlorphenamine 4 mg 3 times daily, vitamin C 0.1g 3 times daily	Xu 2002

Corticosteroids compared to placebo for treating symptomatic, biopsy‐confirmed oral lichen planus
Population: people with oral lichen planus Setting: university dental clinics Intervention: topical corticosteroids Comparison: placebo
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	Number of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Risk with placebo	Risk with corticosteroids
Pain resolution** measured via VAS (1‐10cm) and 5‐grade score Follow‐up: 8‐9 weeks	306 per 1000	584 per 1000  (330 to 1000)	RR 1.91 (1.08 to 3.36)	72  (2 RCTs)	⊕⊕⊝⊝  Low^a,c	1 RCT evaluated clobetasol propionate and the other flucinonide.
Clinical resolution*** measured via Thongparsom and 5‐grade score Follow‐up: 8‐9 weeks	50 per 1000^b	300 per 1000 (38 to 2379)	RR 6.00 (0.76 to 47.58)	72  (2 RCTs)	⊕⊝⊝⊝  Very low^a,d	1 RCT evaluated clobetasol propionate and the other flucinonide.
Adverse effects**** Follow‐up: 3‐9 weeks	89 per 1000	132 per 1000  (43 to 405)	RR 1.48 (0.48 to 4.56)	88  (3 RCTs)	⊕⊝⊝⊝  Very low^a,d	RCTs evaluated clobetasol propionate, flucinonide (no adverse effects noted) and triamcinolone acetonide.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Subjective assessment by participants at the end of treatment Assessment by clinicians at the end of treatment *Reported by participants throughout the study CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio; VAS:** visual analogue scale.
GRADE Working Group grades of evidence  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Corticosteroids compared to calcineurin inhibitors for treating symptomatic, biopsy‐confirmed oral lichen planus
Population: people with oral lichen planus Setting: university dental clinics Intervention: topical corticosteroids Comparison: topical calcineurin inhibitors
Outcomes	Treatment comparison	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
Outcomes	Treatment comparison	Risk with calcineurin inhibitors	Risk with corticosteroids	Relative effect  (95% CI)	№ of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
Pain resolution** measured via VAS and 4‐grade scale Follow‐up: 3‐8 weeks	clobetasol vs ciclosporin	200 per 1000	422 per 1000  (152 to 1000)	RR 2.11 (0.76 to 5.86)	39  (1 RCT)	⊕⊝⊝⊝  Very low^a	—
	clobetasol vs tacrolimus	440 per 1000	198 per 1000  (106 to 387)	RR 0.45 (0.24 to 0.88)	100  (2 RCTs)	⊕⊝⊝⊝  Very low^b	—
	triamcinolone vs tacrolimus	—	—	—	—	—	No data for this outcome
Clinical resolution*** measured via Thongprasom and 4‐grade scale Follow‐up: 3‐8 weeks	clobestol vs ciclosporin	150 per 1000	474 per 1000  (150 to 1000)	RR 3.16 (1.00 to 9.93)	39  (1 RCT)	⊕⊝⊝⊝  Very low^c	—
	clobetasol vs tacrolimus	654 per 1000	399 per 1000  (248 to 647)	RR 0.61 (0.38 to 0.99)	52  (2 RCTs)	⊕⊝⊝⊝  Very low^d	—
	triamcinolone vs tacrolimus	467 per 1000	401 per 1000  (257 to 630)	RR 0.86 (0.55 to 1.35)	60  (2 RCTs)	⊕⊝⊝⊝  Very low^e	—
Adverse effects**** Follow‐up: 3‐8 weeks	clobetasol vs ciclosporin	50 per 1000	316 per 1000  (42 to 1000)	RR 6.32 (0.84 to 47.69)	39  (1 RCT)	⊕⊝⊝⊝  Very low^a	—
	clobetasol vs tacrolimus	180 per 1000	9 per 1000  (0 to 149)	RR 0.05 (0.00 to 0.83)	100  (2 RCTs)	⊕⊝⊝⊝  Very low^f	—
	triamcinolone vs tacrolimus	516 per 1000	243 per 1000  (114 to 511)	RR 0.47 (0.22 to 0.99)	58  (2 RCTs)	⊕⊕⊝⊝  Low^g	—
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Subjective assessment by participants at the end of treatment Assessment by clinicians at the end of treatment * Reported by participants throughout the duration of the study CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio; VAS:** visual analogue scale.
GRADE Working Group grades of evidence  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Date	Event	Description
17 August 2019	New citation required and conclusions have changed	We identified three placebo‐controlled trials (the last version of the review had none) and found low‐certainty evidence that topical corticosteroids may be more effective than placebo for reducing the pain of oral lichen planus. We found very low‐certainty evidence that tacrolimus may reduce pain more than clobetasol, but also that tacrolimus may be more likely to cause adverse effects.
26 February 2019	New search has been performed	Review split into two: one focusing on corticosteroid therapies and one focusing on non‐corticosteroid therapies. Updated searches 14 new studies included

Date	Event	Description
14 June 2011	New citation required and conclusions have changed	Significant update of original review published in 1999: new review team; review methodology updated; included studies increased from 9 to 28.
14 June 2011	New search has been performed	Searches updated to 26 January 2011.
18 June 2008	New search has been performed	Converted to new review format.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Pain score	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 Clobetasol vs placebo	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Pain resolution	2	72	Risk Ratio (M‐H, Fixed, 95% CI)	1.91 [1.08, 3.36]
2.1 Flucinonide vs placebo	1	40	Risk Ratio (M‐H, Fixed, 95% CI)	2.17 [1.03, 4.55]
2.2 Clobetasol vs placebo	1	32	Risk Ratio (M‐H, Fixed, 95% CI)	1.6 [0.67, 3.84]
3 Clinical score	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Clobetasol vs placebo	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Clinical resolution	2	72	Risk Ratio (M‐H, Fixed, 95% CI)	6.0 [0.76, 47.58]
4.1 Flucinonide vs placebo	1	40	Risk Ratio (M‐H, Fixed, 95% CI)	9.0 [0.52, 156.91]
4.2 Clobetasol vs placebo	1	32	Risk Ratio (M‐H, Fixed, 95% CI)	3.0 [0.13, 68.57]
5 Adverse effects	3	88	Risk Ratio (M‐H, Fixed, 95% CI)	1.48 [0.48, 4.56]
5.1 Flucinonide vs placebo	1	40	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 Triamcinolone vs placebo	1	16	Risk Ratio (M‐H, Fixed, 95% CI)	1.29 [0.37, 4.53]
5.3 Clobetasol vs placebo	1	32	Risk Ratio (M‐H, Fixed, 95% CI)	2.0 [0.20, 19.91]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Pain score	4		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
1.1 Clobetasol vs tacrolimus	1	68	Mean Difference (IV, Fixed, 95% CI)	0.08 [‐0.33, 0.49]
1.2 Triamcinolone vs pimecrolimus	2	70	Mean Difference (IV, Fixed, 95% CI)	‐0.05 [‐2.97, 2.87]
1.3 Triamcinolone vs ciclosporin	1	139	Mean Difference (IV, Fixed, 95% CI)	‐0.23 [‐1.02, 0.56]
2 Pain resolution	3		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
2.1 Clobetasol vs ciclosporin	1	39	Risk Ratio (M‐H, Fixed, 95% CI)	2.11 [0.76, 5.86]
2.2 Clobetasol vs tacrolimus	2	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.45 [0.24, 0.88]
3 Clinical score	3		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Clobetasol vs tacrolimus	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Triamcinolone vs pimecrolimus	2		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Clinical resolution	4		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
4.1 Clobetasol vs ciclosporin	1	39	Risk Ratio (M‐H, Fixed, 95% CI)	3.16 [1.00, 9.93]
4.2 Clobetasol vs tacrolimus	2	52	Risk Ratio (M‐H, Fixed, 95% CI)	0.61 [0.38, 0.99]
4.3 Triamcinolone vs tacrolimus	2	60	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.55, 1.35]
5 Adverse effects	7		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
5.1 Clobetasol vs ciclosporin	1	39	Risk Ratio (M‐H, Fixed, 95% CI)	6.32 [0.84, 47.69]
5.2 Clobetasol vs tacrolimus	2	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.05 [0.00, 0.83]
5.3 Triamcinolone vs ciclosporin	1	139	Risk Ratio (M‐H, Fixed, 95% CI)	0.15 [0.05, 0.49]
5.4 Triamcinolone vs pimecrolimus	1	35	Risk Ratio (M‐H, Fixed, 95% CI)	0.21 [0.01, 4.10]
5.5 Triamcinolone vs tacrolimus	2	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.47 [0.22, 0.99]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Pain (mean score)	3	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 Clobetasol + miconazole vs clobetasol + placebo	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Dexamethasone + nystatin + curcumin vs dexamethasone + nystatin + placebo	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.3 Triamcinolone vs triamcinolone + glucosamine sulphate	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Clinical score	3	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Clobetasol + miconazole vs clobetasol + placebo	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Dexamethasone + nystatin + curcumin vs dexamethasone + nystatin + placebo	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Triamcinolone vs triamcinolone + glucosamine sulphate	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Clinical resolution	2	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3.1 Dexamethasone + mycostatin vs dexamethasone alone	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Dexamethasone + nystatin + curcumin vs dexamethasone + nystatin + placebo	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Adverse effects	2	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.1 Clobetasol + miconazole vs clobetasol + placebo	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 Dexamethasone + nystatin + curcumin vs dexamethasone + nystatin + placebo	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	Study design: split‐mouth RCT Conducted in Iran Number of centres: 1 Recruitment period: July 2012 to January 2013 Funding source: supported by Vice Chancellory of Shiraz University of Medical Sciences (no. 90010342)
Participants	Inclusion criteria: bilateral clinically and biopsy confirmed OLP lesions (size > 4 cm), similar in form bilaterally with < 1 cm difference in size Exclusion criteria: systemic disease, pregnancy, drug consumption, smoking, people with lesions contacting dental amalgam, and people with dermal and other mucosal involvement at the time of the therapy. Participants did not receive any treatment for OLP at least 1 month prior to the beginning of the study. Randomised 30 participants; analysed 23 participants at final follow‐up (split‐mouth design)
Interventions	Group A: triamcinolone acetonide ointment 0.1% in Orabase (3 times daily for 1 week, tapered to twice daily on week 2, once daily on week 3, alternate day on week 4 and discontinued at week 5) Group B: single session of cryotherapy with nitrous oxide gas
Outcomes	Pain score (VAS), Thongprasom sign score (based on size and appearance), improvement scale based on EI (no improvement (EI = 0), mild improvement (EI > 0 to 25%), moderate improvement (EI ≥ 25% to < 75%), marked improvement (EI ≥ 75% to < 100%) and healed (EI = 100%) Measured before treatment and after 2, 4 and 6 weeks of treatment
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not feasible.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not feasible.
Incomplete outcome data (attrition bias)   All outcomes	High risk	7 (23%) lost to follow‐up. Quote: "Two patients chose to discontinue the study protocol after 3 weeks and other treatments were begun. Two patients were also excluded on week 4 due to poor compliance and wrong use of ointment. Three patients did not come for the third follow‐up session."
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Iran Number of centres: 1 Recruitment period: October 2012 to June 2013 Funding source: supported by Vice Chancellory of Mashhadd University of Medical Sciences
Participants	Inclusion criteria: people with clinical signs of erosive‐atrophic OLP which was confirmed by clinical and histopathological examination Exclusion criteria: pregnancy; breastfeeding; current use of anticoagulants or antiplatelet agents; current orthodontic treatment; history of gastric ulcers, duodenal ulcers, gallstones, hepatic diseases, any existing malignancy or viral infection in mouth; receiving any topical treatment for OLP in the past 2 weeks or any systemic treatment for OLP in the past 4 weeks; use of azathioprine, ciclosporin or receiving PUVA, UVA or UVB in the last month; history of allergy to corticosteroids or curcumin Group A: randomised 12; analysed 12 Group B: randomised 8; analysed 8
Interventions	Group A: 1 dexamethasone mouthwash 0.5 mg 3 times daily + 1 nystatin (100,000 units) suspension 3 times daily + 4 curcumin tablets (500 mg) twice daily for 4 weeks Group B: 1 dexamethasone mouthwash 0.5 mg 3 times daily + 1 nystatin (100,000 units) suspension 3 times daily + 4 placebo tablets twice daily for 4 weeks
Outcomes	Pain (VAS), clinical score (Thongprasom), complete resolution, adverse effects Measured at baseline, 14 days and 28 days
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization was performed using a computer‐generated random number table."
Allocation concealment (selection bias)	Low risk	Quote: "Randomization was performed using a computer‐generated random number table. Study medication tablets with 95% curcominoids (Samilabs Limited, Bangalore, India) and identical placebo tablets (containing lactose) were prepackaged by a university pharmacist in identical containers. During treatment, both of the practitioners and the patients were unaware of medications they were using."
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "During treatment, both of [sic] the practitioners and the patients were unaware of medications they were using."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "During treatment, both of [sic] the practitioners and the patients were unaware of medications they were using."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in the result analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Italy Number of centres: 1 Recruitment period: unspecified Funding source: supported by the principal investigator's own funds
Participants	Inclusion criteria: histological diagnosis of OLP, on the basis of WHO criteria (Kramer 1978), and presenting painful lesions Exclusion criteria: occurrence dysplasia in the histopathological specimen; use of lichenoid reaction inducing medication and presence of amalgam fillings nearby the lesions; interventions for OLP in the previous 12 weeks; pregnant or breastfeeding women; confirmed or suspected hypersensitivity to any of the chemicals used in the treatment Group A: randomised 18; analysed 16 Group B: randomised 18; analysed 16
Interventions	Group A: clobetasol propionate 0.025% in 4% hydroxyethyl cellulose gel + miconazole gel once daily + 0.12% chlorhexidine mouthrinse twice daily for 8 weeks Group B: hydroxyethyl cellulose twice daily + miconazole gel once daily + 0.12% chlorhexidine mouthrinse twice daily for 8 weeks
Outcomes	Pain (VAS 0–10 cm), clinical score (Thongprasom), adverse effects
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Computer‐generated randomisation list (with blocked randomisation of 4)."
Allocation concealment (selection bias)	Low risk	Quote: "Allocation concealment was ensured by keeping the randomisation list in the care of one of the investigators not involved in the clinical part of the study."
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "During treatment, neither the principal investigator nor the involved subjects knew which one of the two treatments they were using."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "During treatment, neither the principal investigator nor the involved subjects knew which one of the two treatments they were using."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: missing data balanced in numbers across intervention groups (2 and 2) and not likely to have a clinically relevant impact on the intervention effect estimate. Unlikely to have introduced bias.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in India Number of centres: 1 Recruitment period: unspecified Funding source: nil
Participants	Inclusion criteria: people with symptomatic OLP (pain or burning sensation, or both) who were agreeing for the biopsy and were ready to apply the medication supplied Exclusion criteria: people with a history of malignancy, immunocompromised diseases, current systemic or generalised infections, pregnancy or breastfeeding, received topical or systemic immunosuppressants, retinoids or any other systemic therapies known to cause or suspected to have an effect on OLP within the last 4 weeks and allergic to the drugs supplied Group A: randomised 15; analysed 15 Group B: randomised 15; analysed 15
Interventions	Group A: triamcinolone acetonide 0.1% oral paste 4 times daily for 2 months Group B: pimecrolimus 1% oral paste 4 times daily for 2 months
Outcomes	Pain (VAS), clinical score (Thongprasom), presence of erythematous areas
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in USA Number of centres: 1 Recruitment period: February 2003 to August 2004 (then stopped for futility following an interim analysis) Funding source: Mount Zion Health Fund, through UCSF
Participants	Inclusion criteria: aged > 21 years, with a current presentation of atrophic or erosive OLP, and a symptom score for OLP between 3 and 8 at enrolment. All participants had been diagnosed by clinical as well as histopathological examination of oral lesions. Exclusion criteria: pregnancy; breastfeeding; medical contraindication to prednisone or fluconazole; on long‐term corticosteroid therapy; current use of anticoagulants or antiplatelet agents; current orthodontic treatment; history of gastric ulcers, duodenal ulcers, gallstones or liver disease Group A: randomised 17; analysed 17 (ITT analysis) Group B: randomised 16; analysed 16 (ITT analysis)
Interventions	Group A: prednisone 60 mg + curcuminoids 2000 mg daily (in 2 divided doses) for 7 weeks Group B: prednisone 60 mg + placebo daily (in 2 divided doses) for 7 weeks
Outcomes	Pain (VAS), clinical score (modified Oral Mucositis Index), change in symptoms scale, adverse effects
Notes	Among participants who completed the study, all participants in the placebo group had > 85% compliance, and 14 had > 95% compliance (range 87.7–100%); in the curcuminoids group, all but 1 had > 95% compliance (range 64.2–100%). Comment: after enrolling 33 participants, the study was halted for futility following an interim analysis. Estimated post hoc power to detect a difference was 7%.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomisation sequence was generated […] using random number generator."
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "Both patients and investigators were blind."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Both patients and investigators were blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	ITT analysis. 5/33 (15%) participants lost at follow‐up, unbalanced (4 and 1).
Selective reporting (reporting bias)	High risk	Comment: ≥ 1 outcomes of interest were reported incompletely (standard deviations were missing) so they could not be entered in a meta‐analysis.
Other bias	Low risk	5 participants used topical corticosteroids during the study. Of these, 2 were in the placebo group and 3 were in the curcuminoids group (1 of whom dropped out of the study after 4 weeks, at the second follow‐up visit). Unlikely to have caused bias.

Methods	Study design: RCT Conducted in Brazil Number of centres: 1 Recruitment period: February 2012 to November 2012 Funding source: supported by Postgraduate Research Group of Porto Alegre University Hospital
Participants	Inclusion criteria: aged > 21 years, symptomatic atrophic/erosive OLP, histopathological diagnosis of OLP based on WHO‐based criteria Exclusion criteria: pregnancy or nursing women, histological signs of dysplasia, OLP therapy in the previous 3 months, amalgam restorations near the lesions and the use of medications associated with oral lichenoid reactions Group A: randomised 21; analysed 16 Group B: randomised 21; analysed 17
Interventions	Group A: clobetasol propionate gel 0.05% 3 times daily for 4 weeks and nystatin 100,000 UI/mL oral rinses 3 times daily Group B: InGaAIP laser applications, 3 times weekly for 4 weeks (12 sessions) and nystatin 100,000 UI/mL oral rinses 3 times daily
Outcomes	Pain (VAS), clinical score (Thongrasom), functional score (Lilleby score – 5 grades), relapse rate, anxiety questionnaire (Beck Anxiety Inventory), adverse effects Measured once weekly during treatment (days 7, 14, 21 and 30) and 4 weeks (day 60) and 8 weeks (day 90) after treatment
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The patients were randomly assigned to one of the two treatment groups using computer‐generated random number tables."
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "The patient was not blinded to the treatment."
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Comment: participants (self‐assessed, pain was primary outcome) were not blinded.
Incomplete outcome data (attrition bias)   All outcomes	High risk	5/21 (23.8%) participants from Group A and 4/21 (19%) from Group B were lost at follow‐up.
Selective reporting (reporting bias)	High risk	Comment: data on Beck Anxiety Inventory and Lillebay score not reported or impossible to extract.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Egypt Number of centres: 2 (Department of Oral Medicine, Oral Diagnosis and Periodontology, Faculty of Dentistry and the Department of Skin and Venereal Diseases, Faculty of Medicine in Ain Shams University‐Egypt) Recruitment period: unspecified Funding source: supported by the principal investigator's own funds
Participants	Inclusion criteria: systemically free, both genders aged 25–60 years, clinically and histologically confirmed painful erosive or atrophic OLP according to modified WHO criteria Exclusion criteria: history of drug‐induced lichenoid lesions, potential treatment of OLP for < 2 weeks by topical and 4 weeks systemic therapy before study, pregnancy, breastfeeding, smoking and known hypersensitivity or severe adverse effects to the treatment drugs or to any ingredient of their preparation Group A: randomised 15; analysed 15 Group B: randomised 15; analysed 15
Interventions	Group A: betamethasone 17‐valerate 0.1% adhesive gel base, 4 times daily for 4 weeks Group B: pimecrolimus adhesive 1% gel base, 4 times daily for 4 weeks
Outcomes	Pain (VAS), clinical score (5‐grade scale), adverse effects
Notes	All the outcomes were reported in a way that allowed us to perform quantitative analysis.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computer‐generated randomization assignment (blocks of 4)."
Allocation concealment (selection bias)	Low risk	Quote: "Allocation concealment was performed by sealed envelopes containing the randomization number."
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "both patients and outcome assessors were blinded for the type of medication."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "both patients and outcome assessors were blinded for the type of medication."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in the result analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in West China Number of centres: 1 (Department of Oral Medicine, School of Stomatology, Sichuan University) Recruitment period: September 2009 to December 2009 Funding source: unspecified
Participants	Inclusion criteria: histopathologically and clinically diagnosed with erosive OLP; presence of a single erosion lesion; aged from 18 to approximately 70 years; erosive area not exceeding 1.5 cm²; and normal physical examination results before medication (including complete blood cell count, renal and hepatic clinical chemistry examination, urine and stool routine test, blood pressure examination, ultrasonic examination of abdomen, chest x‐ray and electrocardiogram). Exclusion criteria: presence of severe systemic diseases or other severe oral mucous diseases; use of antibiotics within 1 month or immunomodulating drugs within 3 months; history of topical treatment within 1 week; presence of lichenoid reaction caused by amalgam fillings or certain drugs (including beta‐blockers, dapsone, oral hypoglycaemics, non‐steroidal anti‐inflammatory drugs, penicillamine, phenothiazines, sulphonylureas and gold salts); pregnancy, intention of pregnancy, breastfeeding or recent use of steroid hormone‐based contraceptives; history of psychiatric disorders; participation in any other clinical trials in the 3 months before enrolment in study; refusing to follow medical advice; or could not finish the return visits Group A: randomised 20; analysed 19 Group B: randomised 18; analysed 17
Interventions	Group A: amlexanox paste 250 mg, 3 times daily for 1 week Group B: dexamethasone paste 6.45 mg (0.043%), 3 times daily for 1 week
Outcomes	Pain (VAS), size of erosive area, adverse effects
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The patients were assigned to the experimental group or positive‐control group by using a computer generated random number list."
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: the trial authors reported that the containers were the same, but no details were reported about blinding of personnel or the assessor, so we judged the study as unclear.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Comment: the trial authors reported that the containers were the same, but no details were reported about blinding of personnel or assessor, so we judged the study as unclear.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: missing data balanced in numbers across intervention groups (1 and 1) and not likely to have a clinically relevant impact on the intervention effect estimate. Unlikely to have introduced bias.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT (trials authors incorrectly described their study as an "observational prospective study") Conducted in Egypt Number of centres: 1 Recruitment period: June to December 2017 Funding source: unspecified
Participants	Inclusion criteria: any age or sex who had symptomatic OLP (including mixed forms), not related to any identifiable cause (e.g. lichenoid drug reactions), and confirmed by histopathology Exclusion criteria: people undergoing systemic chemotherapy, people who were immunosuppressed or had malignant diseases, pregnant women, and history of hypersensitivity to hyaluronic acid Group A: randomised 20; analysed 20 Group B: randomised 20; analysed 20
Interventions	Group A: triamcinolone preparation 0.1%, 3 times daily for 28 days Group B: hyaluronic acid preparation 0.2%, 3 times daily for 28 days
Outcomes	Pain (VAS), clinical score (modified oral mucositis index), size of erosive areas
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in the result analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: 2‐arm cross‐over RCT Conducted in UK Number of centres: 1 Recruitment period: unclear Funding source: unspecified
Participants	Inclusion criteria: symptomatic erosive or ulcerative OLP, had not previously used fluticasone or betamethasone, and were generally healthy Exclusion criteria: unclear Group A: randomised 22; analysed 19 Group B: randomised 22; analysed 20
Interventions	Fluticasone 50 μg spray, 2 puffs applied to lesions 4 times daily, for 6 weeks and betamethasone sodium phosphate 500 μg (0.5 mg in 10 mL of water), 3 minutes mouthrinse 4 times daily for 6 weeks Group A: sequence fluticasone and betamethasone Group B: sequence betamethasone and fluticasone Washout period: 2 weeks
Outcomes	Pain (VAS), clinical score (modified from Thongprasom), size of affected mucosa, quality of life (OHIP, OHQoL), adverse effects
Notes	All 5 people who left the study (3 in group A and 2 in group B) did so during fluticasone treatment because of adverse effects.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "A table of random numbers was used."
Allocation concealment (selection bias)	Low risk	Quote: "The sequence was concealed until the effect of both intervention was analysed."
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not feasible.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not feasible.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: missing data balanced in numbers across intervention groups (3 and 2) and not likely to have a clinically relevant impact on the intervention effect estimate. Unlikely to have introduced bias.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported. Comment: data from this cross‐over study were not analysed as paired and, therefore, they could not be entered in a meta‐analysis.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: 3‐arm RCT Conducted in Egypt Number of centres: 2 Recruitment period: unclear Funding source: unspecified
Participants	Inclusion criteria: both sexes, aged 25–60 years, with clinically and histologically confirmed erosive/atrophic OLP and free from any systemic disease Exclusion criteria: people with history of drug‐induced lichenoid lesions or potential treatment for OLP (for < 2 weeks by topical or 4 weeks by systemic therapy) before start of study; pregnant or breastfeeding women; smokers; known hypersensitivity or adverse effects to treatment drugs; losses of pliability or flexibility in tissues involved by oral lesions or histological signs of epithelial dysplasia or lichenoid lesions Group A: randomised 10; analysed 10 Group B: randomised 10; analysed 10 Group C: randomised 10; analysed 10
Interventions	Group A: topical triamcinolone acetonide 0.1% oral paste (GEO ORALOG; Geopharma, Cairo, Egypt) 4 times daily + glucosamine sulphate 500 mg capsules orally 3 times daily (glucosamine; GlaxoSmithKline, Cairo, Egypt) for 8 weeks Group B: topical triamcinolone acetonide 0.1% oral paste twice daily + glucosamine sulphate 500 mg 3 times daily for 8 weeks  Group C: topical triamcinolone acetonide 0.1% oral paste alone 4 times daily for 8 weeks
Outcomes	Pain (VAS, resolution), clinical score (Thongprasom Classification score) Measured at baseline, 1, 2, 4, 6, 8 and 12 weeks
Notes	We compared the outcomes of Group A and Group C as the triamcinolone daily dose was the same (4 times daily).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomised into three groups (10 patients each) using computer‐generated randomisation assignment."
Allocation concealment (selection bias)	Low risk	Quote: "allocation concealment with sealed envelopes."
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Comment: participants (self‐assessed, pain was primary outcome) were not blinded.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in result analysis.
Selective reporting (reporting bias)	High risk	Adverse effects, mentioned in the Methods were not reported in the Results.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Sri Lanka Number of centres: 1 (Oral Medicine Clinic, University Dental Hospital, Peradeniya, Sri Lanka) Recruitment period: June 2014 to July 2015 Funding source: University Research Grant RG/2013/08/D
Participants	Inclusion criteria: histologically confirmed symptomatic OLP with main involvement of the bilateral buccal mucosae Exclusion criteria: taken any topical or systemic medication for OLP in the previous 3 months; contraindication for use of medications, including history of allergy to either corticosteroids or tacrolimus; history suggestive of immunosuppression; history of developing possible lichenoid reactions either drug induced or due to dental amalgam; or pregnant or breastfeeding; people with diabetes taking oral hypoglycaemic drugs and people with concurrent skin/genital LP lesions. Group A: randomised 34; analysed 34 Group B: randomised 34; analysed 34
Interventions	Group A: clobetasol 0.05% cream, twice daily for 3 weeks Group B: tacrolimus 0.1% cream, twice daily for 3 weeks
Outcomes	Pain (VAS, resolution), clinical score (Thongprasom Classification score), adverse effects Measured at baseline and 3 weeks
Notes	Pain and clinical score were recorded separately for right and left mucosal surfaces. As pooling was not feasible, we decided to randomly select 1 side only (right)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomisation sequence for patients was sourced using computer‐generated numbers."
Allocation concealment (selection bias)	Low risk	Quote: "that were sent in sealed envelopes to a trained nurse who was responsible for giving the assigned treatment drug to each patient."
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "Medications were identically prepackaged to maintain examiner and patient blinding."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Medications were identically prepackaged to maintain examiner and patient blinding."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data. All randomised participants included in analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Iran, Guilian University of Medical Sciences Number of centres: 1 Recruitment period: 2 months Funding source: unspecified
Participants	Inclusion criteria: clinical signs of OLP (atrophic and ulcerative forms), confirmed by clinical and histopathological examination Exclusion criteria: pregnancy and breastfeeding; current use of anticoagulants or antiplatelet agents; current orthodontic treatment; history of gastric ulcers, duodenal ulcers, gallstones, hepatic diseases, any existing malignancy or viral infections in the mouth; history of topical treatment for OLP in the preceding 2 weeks or any systemic treatment for OLP in the preceding 4 weeks; taking azathioprine, ciclosporin or receiving Psoralen + ultraviolet A, ultraviolet A or ultraviolet B radiation in the past month; history of allergy to corticosteroids or curcumin Group A: randomised 25; analysed 25 Group B: randomised 25; analysed 25
Interventions	Group A: curcumin paste 5%, 3 times daily for 4 weeks Group B: triamcinolone acetonide paste 0.1%, 3 times daily for 4 weeks
Outcomes	Complete remission (100% reduction of signs and symptoms), pain (VAS), clinical response (Thongprasom Classification score) Appearance score and severity of pain assessed at baseline and at end of 2 and 4 weeks
Notes	Trial registered in an unspecified register with code number IRCT2001105012950N2 Authors reported in the Results section that 5/55 participants were lost at follow‐up, while in the rest of the report they stated that the trial enrolled 50 participants and that all of them were included in the analysis.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The sample size of 25 was chosen based on expected and actual enrolment of study subjects over a two‐month time period. A blocked randomisation (block size of six) was used. The pharmacy of Guilan University of Medical Sciences (…) generated the randomisation sequence using the random number generator in Microsoft Excel (Microsoft Corp, Seattle, WA)."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Quote: "Both participants and investigators were blinded to the treatment assignment." However, one intervention is yellow and the other white, so we assessed this as unclear.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Both participants and investigators were blinded to the treatment assignment."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data. All randomised participants included in result analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

PERMALINK

Interventions for treating oral lichen planus: corticosteroid therapies

Giovanni Lodi

Maddalena Manfredi

Valeria Mercadante

Ruth Murphy

Marco Carrozzo

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

'Summary of findings' table

Results

Description of studies

Results of the search

1.

Included studies

Characteristics of the trials

Design

Setting

Funding

Characteristics of the interventions

1. Characteristics of the interventions.

Characteristics of the outcomes

Pain

Clinical response

Adverse effects

Other outcomes not relevant for this review

Excluded studies

Studies awaiting classification

Ongoing studies

Risk of bias in included studies

Allocation

Blinding

Performance bias

Detection bias

Incomplete outcome data

Selective reporting

Other potential sources of bias

Overall risk of bias

2.

3.

Effects of interventions

Summary of findings for the main comparison. Corticosteroids compared to placebo for treating oral lichen planus.

Summary of findings 2. Corticosteroids versus calcineurin inhibitors for treating oral lichen planus.

Methods	Study design: RCT Conducted in Netherlands Number of centres: 1 Recruitment period: 2001–2004 Funding source: unspecified
Participants	Inclusion criteria: Caucasian people with a confirmed diagnosis of symptomatic OLP based on clinical and histopathological features Exclusion criteria: aged < 18 years; histopathological examination with atypical or lichenoid dysplastic features; asymptomatic oral lesions and specific treatment within 4 weeks prior to the study Group A: randomised 20; analysed 20 Group B: randomised 20; analysed 20
Interventions	Group A: tacrolimus 0.1% ointment, 4 times daily for 6 weeks Group B: triamcinolone acetonide 0.1% in hypromellose 20% ointment, 4 times daily for 6 weeks
Outcomes	Clinical score (4‐grade score), adverse effects
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "pre‐determined randomisation list stratified by sex."
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in: Egypt Number of centres: 1 Recruitment period: unclear Funding source: unspecified
Participants	Inclusion criteria: oral erosive lesions were diagnosed according to Andreason classification; histological confirmation of OLP according to the WHO's clinicopathological diagnostic criteria for OLP; no previous treatment of OLP for ≥ 3 months; willingness and ability to complete the present clinical trial; aged > 35 years without skin involvement Exclusion criteria: histological signs of dysplasia; using drugs associated with lichenoid reaction; pregnant or breastfeeding; smoker; systemic diseases such as immunodysfunction, haematological and hepatological; had photosensitivity history Group A: randomised 10; analysed 10 Group B: randomised 10; analysed 10
Interventions	Group A: triamcinolone acetonide 0.1% in Orabase (Kenacort) 3 times daily for 6 weeks Group B: photodynamic therapy mediated by methylene blue once weekly for 2 months
Outcomes	Pain (VAS), clinical score (Thongprasom)
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not feasible.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not feasible.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: no missing data. All randomised participants included in the study's analyses.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Iran (Department of Ora Medicine, Mashhad Dental Faculty) Number of centres: 1 Recruitment period: 2008–2010 Funding source: Vice Chancellor for Research, Mashhad University of Medical Sciences
Participants	Inclusion criteria: adults with atrophic‐erosive OLP as confirmed by biopsy, lesion size < 2 cm; atrophic‐erosive lesions limited to 2 sides of the oral cavity; no current treatment with immunomodulatory agents; a 1‐week washout period was required prior to enrolment if participants were taking immunomodulatory agents. Exclusion criteria: inability to undergo oral biopsy for diagnosis; aged < 18 years; systemic diseases or malignancy, pregnancy, lesion/lesions with dysplasia; history of allergic reaction to corticosteroids or immunomodulatory drugs; presence of lesions adjacent to an amalgam filling Group A: randomised 14; analysed 6 Group B: randomised 14; analysed 4
Interventions	Group A: pimecrolimus 1% cream, 3 times daily for 2 months Group B: triamcinolone acetonide 0.1% in Orabase, 3 times daily for 2 months
Outcomes	Pain (VAS), clinical score (Thongprasom), improvement of lesions
Notes	Reporting of VAS and clinical score values did not allow quantitative analysis.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomly divided in two groups, based on a random numbering table."
Allocation concealment (selection bias)	Low risk	Quote unpublished: "We used sealed envelopes for each patient for guarantee allocation concealment."
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "Patients were aware of the medication allocated."
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Comment: participants (self‐assessed, pain was primary outcome) were not blinded.
Incomplete outcome data (attrition bias)   All outcomes	High risk	8/14 (57%) participants from Group A and 10/14 (71%) participants from Group B were lost at the 4th and 5th visit.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in: Sweden Number of centres: 1 Recruitment period: unspecified Funding source: unspecified
Participants	Inclusion criteria: consecutive patients with clinical and histological diagnosis of OLP according to WHO recommendations Exclusion criteria: lesions adjacent to buccal or lingual amalgam fillings Group A: randomised 20; analysed 17 Group B: randomised 20; analysed 20
Interventions	Group A: clobetasol propionate 0.025% ointment in Orabase (50/50) twice daily for first 3 weeks, once daily for second 3 weeks, once every other day for third 3 weeks (total 9 weeks) Group B: triamcinolone acetonide 0.1% ointment in Orabase (50/50) twice daily for first 3 weeks, once daily for second 3 weeks, once every other day for third 3 weeks (total 9 weeks)
Outcomes	Pain (VAS), clinical score (4‐grade score), adverse effects
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: "randomised" – insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "Double blind."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Double blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Comment: missing data were not likely to have a clinically relevant impact on the intervention effect estimate (3/40 (7.5%)). It must be noted that lost participants were all from the clobetasol group. Unlikely to have introduced bias.
Selective reporting (reporting bias)	High risk	Comment: ≥ 1 outcomes of interest were reported incompletely (standard deviations were missing) so data could not be entered in a meta‐analysis.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: 3‐arm parallel RCT Conducted in Finland Number of centres: 2 (Oulu and Kuopio) Recruitment period: June 2014 to December 2014 Funding source: Finnish Dental Society Apollonia, EVO Funds Oulu University Hospital, EVO/VTR Funds Kuopio University Hospital, Sigrid Juselius Foundation, and MRC Oulu University Hospital
Participants	Inclusion criteria: diagnosis of symptomatic OLP, clinical score ≥ 20 (including VAS > 0), aged > 18 years, washout period of 2 weeks Exclusion criteria: pregnancy or current nursing; allergy to TAC, other macrolides or other substances used in the study medications; hepatic insufficiency; and use of medications that could have significant interactions with TAC, including ciclosporin, erythromycin, rifamycin, posaconazole, itraconazole, ketoconazole, fluconazole, voriconazole, rifampicin, phenytoin and dabigatran Group A: randomised 11; analysed 11 Group B: randomised 7; analysed 7 Group C: randomised 9; analysed 9
Interventions	Group A: tacrolimus ointment 0.1% 3 times daily for 3 weeks Group B: triamcinolone acetonide ointment 0.1% 3 times daily for 3 weeks Group C: placebo (Orabase) 3 times daily for 3 weeks
Outcomes	Changes in clinical scores (modified from Setterfield) from baseline to week 3, changes in VAS scores from baseline to week 3, adverse events
Notes	Since clinical score and pain were reported only as percentage of change from baseline, they were not included in the quantitative analysis.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Computer‐generated randomisation lists."
Allocation concealment (selection bias)	Low risk	Quote: "Allocation concealment was ensured by keeping the randomisation lists in the care of one of the investigators (TS) who was not involved in the clinical part of the study."
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "double blind."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "double blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	ITT analysis. 4/27 (15%) participants lost at follow‐up, balanced across groups (2, 1 and 1).
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: 3‐arm RCT Conducted in: India Number of centres: 1 (Department of Oral Medicine and Radiology, Rajah Muthiah Dental College and Hospital, Annamalai University) Recruitment period: September 2012 to July 2013 Funding source: unspecified
Participants	Inclusion criteria: clinically symptomatic oral lesions confirmed by histologically to be OLP Exclusion criteria: people who had undergone treatment for OLP within 4 weeks of study, pregnant or nursing women, skin lesions, and lesion extending to soft palate and tonsils Group A: randomised 10; analysed 10 Group B: randomised 10; analysed 10 Group C: randomised 10; analysed 10
Interventions	Group A: clobetasol propionate 0.05% topical 4 times daily for 6 weeks Group B: triamcinolone acetonide 0.1% topical 4 times daily for 6 weeks Group C: tacrolimus 0.03% topical 4 times daily for 6 weeks
Outcomes	Complete resolution after 6 weeks
Notes	Complete resolution: clinical resolution or having reticular lesions only
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "using randomisation chart."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Insufficient information to permit judgement, apart from the title of the paper: "triple‐blind."
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Insufficient information to permit judgement, apart from the title of the paper: "triple‐blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data. All randomised participants included in the analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: 3‐arm RCT Conducted in India Number of centres: 1 Recruitment period: December 2013 to August 2015 Funding source: unspecified
Participants	Inclusion criteria: clinically and histopathologically confirmed OLP without dysplasia in histopathological evaluation and willing to take part in study. Clinical diagnosis was based on the presence of interlacing white striations with intermixed erythematous or ulcerative areas. Histopathological diagnosis based on presence of hydropic degeneration of basal cell layer, dense subepithelial inflammatory infiltrate; participants with symptomatic OLP, i.e. burning sensation, and who had not undergone any previous treatment for the same in the last 6 months. Exclusion criteria: evidence of lichenoid reaction in clinical or histopathological assessment; having extra oral manifestations of OLP; long‐term glucocorticosteroid therapy for other systemic diseases; with other white lesions such as leukoplakia; systemic lupus erythematosus along with OLP; pregnancy, breastfeeding; history of allergic reactions to corticosteroids or herbal preparations Group A: randomised 25; analysed 25 Group B: randomised 25; analysed 19 Group C: randomised 25; analysed 25
Interventions	Group A: triamcinolone acetonide 0.1% 3 times daily for 2 weeks Group B: Curenext Oral Gel (Piramel, Health Care, India each gram of which contains curcuma longa extracts 10 mg) 3 times daily for 2 weeks Group C: Curenext Oral Gel (Piramel, Health Care, India each gram of which contains curcuma longa extracts 10 mg) 6 times daily for 2 weeks
Outcomes	Pain (numerical rating scale), clinical score (Modified Oral Mucositis Index)
Notes	We compared the outcomes of group A and group C (highest Curenext Oral Gel dose).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not feasible.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not feasible.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data in the 2 groups considered for analysis. All randomised participants included in the result analysis.
Selective reporting (reporting bias)	Low risk	Planned outcomes reported.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in: The Netherlands Number of centres: 1 Recruitment period: unspecified Funding source: company provided drugs
Participants	Inclusion criteria: clinical features, history, histopathological and immunofluorescence microscopy, non‐use of medication and the absence of oral mucosal lesions other than LP Exclusion criteria: unspecified Group A: randomised 20; analysed 20 Group B: randomised 20; analysed 20
Interventions	Group A: fluocinonide 0.025% in 40% hypromellose ointment in white soft paraffin least 6 times daily for 9 weeks Group B: placebo at least 6 times daily for 9 weeks
Outcomes	Pain (5‐grade score: complete, good, partial, no effect, increase), clinical score (5‐grade score: complete, good, partial, no effect, increase)
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomised." Insufficient information to permit judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "Double blind."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Double blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data. All randomised participants included in the result analysis.
Selective reporting (reporting bias)	Low risk	No other sources of bias identified.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in China Number of centres: 1 Recruitment period: unspecified Funding source: unspecified
Participants	Inclusion criteria: clinical demonstration of hyperaemic OLP, pathological examination confirmed the diagnosis of OLP, history of OLP for ≥ 3 months Exclusion criteria: OLP did not locate in the buccal mucosa, with other oral mucosa diseases or systematic diseases, first‐onset OLP or history < 3 months, already taking local or systematic therapy in the past 3 months, OLP after the treatment of orthodontics and metallic full crown Group A: randomised 37; analysed 37 Group B: randomised 20; analysed 20
Interventions	Group A: alternative use of mycostatin paste and dexamethasone paste (twice daily each) for 6 weeks Group B: dexamethasone paste (4 times daily) for 6 weeks
Outcomes	Clinical score (4‐grade score)
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of sequence generation not described by translator.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: "Double blind."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "Double blind."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No missing data. All randomised participants included in analysis.
Selective reporting (reporting bias)	Unclear risk	Insufficient information to permit judgement.
Other bias	Low risk	No other sources of bias identified.

Methods	Study design: RCT Conducted in Singapore, South Korea, India, Thailand Number of centres: 4: Singapore, Seoul (South Korea), Madras (India), Bangkok (Thailand) Recruitment period: 2000–2003 Funding source: National Medical Research Council, Ministry of Health, Singapore. Trident Pharm Pte Ltd (Singapore) assisted with the distribution of Kenalog to the overseas centres; Novartis Pte Ltd (Singapore) provided Sandimmun Neoral (ciclosporin) at cost for the trial and assisted in co‐ordinating the trial supplies at the overseas centres
Participants	Inclusion criteria: histologically confirmed OLP and with both clinical signs and symptoms of OLP requiring treatment Exclusion criteria: treated previously by either of the trial medications and had worsened during that treatment or had uncontrolled or severe hypertension, serious active or recurrent infections, severe respiratory, renal, or heart disease, recent history of malignancy, insulin‐dependent diabetes, active peptic ulcer disease, active inflammatory gastrointestinal disease or pregnancy Group A: randomised 71; analysed 71 Group B: randomised 68; analysed 68
Interventions	Group A: ciclosporin solution 0.1%, 3 times daily for 8 weeks Group B: triamcinolone acetonide 0.1% in Orabase, 3 times daily for 8 weeks
Outcomes	Pain (VAS), clinical score (Thongprasom)
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "patients were randomly assigned to receive steroid or cyclosporin through central randomization office of the Clinical trials and Epidemiology Research Unit, Singapore."
Allocation concealment (selection bias)	Low risk	Quote: "Once eligibility had been confirmed […] patients were randomly assigned […] by telephone (Singapore and South Korea) or sealed envelope (India and Thailand)."
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Comment: insufficient information to permit judgement.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	ITT analysis. 23/139 (16.5%) participants lost at follow‐up, balanced (13 and 10).
Selective reporting (reporting bias)	Low risk	Clinical improvement, pain and adverse effects reported.
Other bias	Low risk	No other sources of bias identified.