Interventions for managing oral submucous fibrosis

Abstract

Background

Oral submucous fibrosis (OSF) is a chronic disease of the oral cavity that causes progressive constriction of the cheeks and mouth accompanied by severe pain and reduced mouth opening. OSF has a significant impact on eating and swallowing, affecting quality of life. There is an increased risk of oral malignancy in people with OSF. The main risk factor for OSF is areca nut chewing, and the mainstay of treatment has been behavioural interventions to support habit cessation. This review is an update of a version last published in 2008.

Objectives

To evaluate the benefits and harms of interventions for the management of oral submucous fibrosis.

Search methods

We used standard, extensive Cochrane search methods. The latest search date was 5 September 2022.

Selection criteria

We considered randomised controlled trials (RCTs) of adults with a biopsy‐confirmed diagnosis of OSF treated with systemic, locally delivered or topical drugs at any dosage, duration or delivery method compared against placebo or each other. We considered surgical procedures compared against other treatments or no active intervention. We also considered other interventions such as physiotherapy, ultrasound or alternative therapies.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were 1. participant‐reported resumption of normal eating, chewing and speech; 2. change or improvement in maximal mouth opening (interincisal distance); 3. improvement in range of jaw movement; 4. change in severity of oral/mucosal burning pain/sensation; 5. adverse effects. Our secondary outcomes were 6. quality of life; 7. postoperative discomfort or pain as a result of the intervention; 8. participant satisfaction; 9. hospital admission; 10. direct costs of medication, hospital bed days and any associated inpatient costs for the surgical interventions. We used GRADE to assess certainty of evidence for each outcome.

Main results

We included 30 RCTs (2176 participants) in this updated review. We assessed one study at low risk of bias, five studies at unclear risk of bias and 24 studies at high risk of bias.

We found diverse interventions, which we categorised according to putative mechanism of action. We present below our main findings for the comparison 'any intervention compared with placebo or no active treatment' (though most trials included habit cessation for all participants). Results for head‐to‐head comparisons of active interventions are presented in full in the main review.

Any intervention versus placebo or no active treatment

Participant‐reported resumption of normal eating, chewing and speech

No studies reported this outcome.

Interincisal distance

Antioxidants may increase mouth opening (indicated by interincisal distance (mm)) when measured at less than three months (mean difference (MD) 3.11 mm, 95% confidence interval (CI) 0.46 to 5.77; 2 studies, 520 participants; low‐certainty evidence), and probably increase mouth opening slightly at three to six months (MD 8.83 mm, 95% CI 8.22 to 9.45; 3 studies, 620 participants; moderate‐certainty evidence). Antioxidants may make no difference to interincisal distance at six‐month follow‐up or greater (MD −1.41 mm, 95% CI −5.74 to 2.92; 1 study, 90 participants; low‐certainty evidence).

Pentoxifylline may increase mouth opening slightly (MD 1.80 mm, 95% CI 1.02 to 2.58; 1 study, 106 participants; low‐certainty evidence).

However, it should be noted that these results are all less than 10 mm, which could be considered the minimal change that is meaningful to someone with oral submucous fibrosis.

The evidence was very uncertain for all other interventions compared to placebo or no active treatment (intralesional dexamethasone injections, pentoxifylline, hydrocortisone plus hyaluronidase, physiotherapy).

Burning sensation

Antioxidants probably reduce burning sensation visual analogue scale (VAS) scores at less than three months (MD −30.92 mm, 95% CI −31.57 to −30.27; 1 study, 400 participants; moderate‐certainty evidence), at three to six months (MD −70.82 mm, 95% CI −94.39 to −47.25; 2 studies, 500 participants; moderate‐certainty evidence) and at more than six months (MD −27.60 mm, 95% CI −36.21 to −18.99; 1 study, 90 participants; moderate‐certainty evidence).

The evidence was very uncertain for the other interventions that were compared to placebo and measured burning sensation (intralesional dexamethasone, vasodilators).

Adverse effects

Fifteen studies reported adverse effects as an outcome. Six of these studies found no adverse effects. One study evaluating abdominal dermal fat graft reported serious adverse effects resulting in prolonged hospital stay for 3/30 participants. There were mild and transient general adverse effects to systemic drugs, such as dyspepsia, abdominal pain and bloating, gastritis and nausea, in studies evaluating vasodilators and antioxidants in particular.

Authors' conclusions

We found moderate‐certainty evidence that antioxidants administered systemically probably improve mouth opening slightly at three to six months and improve burning sensation VAS scores up to and beyond six months. We found only low/very low‐certainty evidence for all other comparisons and outcomes. There was insufficient evidence to make an informed judgement about potential adverse effects associated with any of these treatments. There was insufficient evidence to support or refute the effectiveness of the other interventions tested.

High‐quality, adequately powered intervention trials with a low risk of bias that compare biologically plausible treatments for OSF are needed. It is important that relevant participant‐reported outcomes are evaluated.

Plain language summary

Interventions for the management of oral submucous fibrosis

Review question

Which treatments are effective in improving the symptoms associated with oral submucous fibrosis?

Key messages

– The overall results are mixed but indicate that using antioxidant medications may be useful to treat restricted mouth opening and are likely to improve the burning sensation in the mouth that is experienced by people who have oral submucous fibrosis.

What is oral submucous fibrosis?

Oral submucous fibrosis is a disease that causes increasing tightness of the cheeks and mouth. People with this condition often have persistent burning mouth pain. These problems can make eating, speaking and swallowing more difficult. Many medicines have been suggested to manage this condition and may be taken by mouth (systemically), applied locally to the surface (topically) or injected directly into the affected areas. Different forms of surgery or physiotherapy are also available.

What did we want to know?

We wanted to find out which treatments are effective for improving the symptoms of oral submucous fibrosis, and which treatments are the most effective. We also wanted to know what the risks and side effects of each treatment might be and how common these are.

What did we do?

We searched databases of medical and dental journals and research trials. We only selected trials known as randomised controlled trials. In this type of trial, participants are allocated to groups randomly. One group receives the intervention and the other receives a different treatment or no treatment at all. These trials aim to reduce the risk of introducing bias in clinical trials.

We wanted to focus on how well treatments worked to allow people with oral submucous fibrosis to return to normal eating, chewing and speaking. Because no trials looked at these measures, we chose increase in mouth opening (measured in millimetres between the upper and lower front teeth) and reduction oral burning sensation (measured on a scale from 0 to 100) as the most likely measures to improve quality of life in people with oral submucous fibrosis. We looked for details of these measures immediately after treatment (up to three months), in the medium term (three to six months) and in the longer term (after six months). We also looked for information on any 'adverse effects' (negative side effects of the treatments).

What did we find?

We found 30 relevant trials. Most trials looked at different treatments that each worked differently. Many trials compared different types of treatments with each other. We decided that trials comparing one treatment against no active treatment were most important because we do not know which, if any, treatments actually work.

The results told us that antioxidant medicines (which scavenge and neutralise unstable particles that are naturally formed during metabolism or by exposure to environmental toxins) taken by mouth may improve mouth opening slightly for up to six months after treatment, but we are not sure that this lasts longer than six months. Antioxidants may also reduce burning sensation in the mouth for more than six months after treatment and we think this might be an important improvement for people with oral submucous fibrosis.

The evidence for other treatments that we looked at was very uncertain.

Only half of the trials reported side effects or other potentially harmful effects of the treatments, so it is difficult to say confidently how safe any of the treatments might be.

What are the limitations of the evidence?

We are moderately confident that antioxidants help in the treatment of oral submucous fibrosis, but have little confidence in the other evidence because many trials had design limitations.

How up to date is the search?

We searched for trials up to 5 September 2022.

Summary of findings

Summary of findings 1. Any intervention versus no active treatment or placebo for managing oral submucous fibrosis.

Any intervention versus no active treatment or placebo for managing oral submucous fibrosis
Patient: people with oral submucous fibrosis Setting: any primary, secondary or tertiary care setting Intervention: any medical, surgical or physical therapy Comparison: placebo or no active treatment
Outcomes	Follow‐up	Illustrative comparative risks (95% CI)		Number of participants (number of studies)	Certainty of the evidence (GRADE)	Comments
		Assumed risk	Corresponding risk
Participant‐reported resumption of normal eating, chewing and speech	Outcome not reported by included studies
Antioxidant vs placebo
Interincisal distance (mm) Increased value = improvement	< 3 months	Mean interincisal distance for placebo groups was 24.50	MD 3.11 increase (0.46 increase to 5.77 increase)	520 (2)	⊕⊕⊖⊖ Lowa,b	Favours antioxidant
	3–6 months	Mean interincisal distance for placebo groups was 25.56	MD 8.83 increase (8.22 increase to 9.45 increase)	620 (3)	⊕⊕⊕⊖ Moderateb	Favours antioxidant
	> 6 months	Mean interincisal distance for placebo group was 30.37	MD 1.41 reduction (5.74 reduction to 2.92 increase)	90 (1)	⊕⊕⊖⊖ Lowc	—
Burning sensation (VAS, 0–100) Reduced value = improvement	< 3 months	Mean VAS for placebo groups was 71.05	MD 30.92 reduction (31.57 reduction to 30.27 reduction)	400 (1)	⊕⊕⊕⊖ Moderateb	Favours antioxidant
	3–6 months	Mean VAS for placebo groups was 72.74	MD 70.82 reduction (94.39 reduction to 47.25 reduction)	500 (2)	⊕⊕⊕⊖ Moderateb,e	Favours antioxidant
	> 6 months	Mean VAS for placebo group was 42.3	MD 27.60 reduction (36.21 reduction to 18.99 reduction)	90 (1)	⊕⊕⊕⊖ Moderated	Favours antioxidant
Adverse effects	Any	0	2 (minor) – flatulence and bloating	90 (1)	—	—
Pentoxifylline vs placebo
Interincisal distance (mm) Increased value = improvement	< 3 months	Mean interincisal distance for placebo group was 23.10	MD 1.80 increase (1.02 increase to 2.58 increase)	106 (1)	⊕⊕⊖⊖ Lowb,d	Favours pentoxifylline
	3–6 months	Mean interincisal distance for placebo groups was 24.66	MD 0.15 reduction (5.04 reduction to 4.74 increase)	136 (2)	⊕⊖⊖⊖ Very lowa,b,d	—
	> 6 months	—	—	—	—	No data for this outcome
Burning sensation (VAS, 0–100) Reduced value = improvement	< 3 months	—	—	—	—	No data for this outcome
	3–6 months	Mean VAS for placebo group was 33	MD 7.00 reduction (28.8 reduction to 42.8 increase)	30 (1)	⊕⊖⊖⊖ Very lowb,c	—
	> 6 months	—	—	—	—	No data for this outcome
Adverse effects	No adverse effects data reported
Dexamethasone vs placebo
Interincisal distance (mm) Increased value = improvement	< 3 months	Mean interincisal distance for placebo group was 28.50	MD 1.80 reduction (6.30 reduction to 2.70 increase)	25 (1)	⊕⊖⊖⊖ Very lowb,c	—
	3–6 months	Mean interincisal distance for placebo group was 29.00	MD 2.1 reduction (6.77 reduction to 2.57 increase)	25 (1)	⊕⊖⊖⊖ Very lowb,c	—
	> 6 months	—	—	—	—	No data for this outcome
Burning sensation (VAS, 0–100) Reduced value = improvement	< 3 months	Mean VAS for placebo group was 52.00	MD 52.00 reduction (62.04 reduction to 41.96 reduction)	25 (1)	⊕⊖⊖⊖ Very lowb,c	—
	3–6 months	Mean VAS for placebo group was 46.00	MD 46.00 reduction (53.44 reduction to 38.56 reduction)	25 (1)	⊕⊖⊖⊖ Very lowb,c	—
	> 6 months	—	—	—	—	No data for this outcome
Adverse effects	Any	No adverse effects		25 (1)	—	—
Vasodilator vs placebo
Interincisal distance (mm) Increased value = improvement	< 3 months' follow‐up	Mean interincisal distance for placebo groups was 28.51	MD 2.51 increase (0.99 reduction to 6.01 increase)	85 (2)	⊕⊖⊖⊖ Very lowb,c	—
	3–6 months	Mean interincisal distance for placebo groups was 29.30	MD 4.04 increase (1.93 reduction to 10.02 increase)	85 (2)	⊕⊖⊖⊖ Very lowb,c	—
	> 6 months	—	—	—	—	No data for this outcome
Burning sensation (VAS, 0–100) Reduced value = improvement	< 3 months	Mean VAS for placebo groups was 64.75	MD 38.79 reduction (52.06 reduction to 25.53 reduction)	85 (2)	⊕⊖⊖⊖ Very lowa,b,d	—
	3–6 months	Mean VAS for placebo groups was 60.48	MD 51.02 reduction (72.85 reduction to 29.20 reduction)	85 (2)	⊕⊖⊖⊖ Very lowa,b,d	—
	> 6 months	—	—	—	—	No data for this outcome
Adverse effects	Any	0	14 (minor) – transient flushing and upper trunk warmness	85 (2)	—	—
Hydrocortisone + hyaluronidase vs placebo
Interincisal distance (mm) Increased value = improvement	< 3 months	—	—	—	—	No data for this outcome
	3–6 months	Mean interincisal distance for placebo group was 33.13	MD 4.88 reduction (12.87 reduction to 3.11 increase)	12 (1)	⊕⊖⊖⊖ Very lowb,c
	> 6 months	—	—	—	—	No data for this outcome
Burning sensatione (VAS, 0–100) Reduced value = improvement	—	—	—	—	—	No data for this outcome
Adverse effects	Any	No adverse effects data reported		N/A	—	—
Physiotherapy vs no active treatment
Interincisal distance (mm) Increased value = improvement	< 3 months	—	—	—	—	No data for this outcome
	3–6 months	Mean interincisal distance for control group was 33.13	MD 0.50 increase (7.72 reduction to 8.72 increase)	24 (1)	⊕⊖⊖⊖ Very lowb,c
	> 6 months	—	—	—	—	No data for this outcome
Burning sensatione (VAS, 0–100) Reduced value = improvement	—	—	—	—	—	No data for this outcome
Adverse effects	No adverse effects data reported
CI: confidence interval; MD: mean difference; 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.

^a Downgraded one level due to inconsistency (substantial heterogeneity).
^b Downgraded one level due to serious concerns regarding risk of bias.
^c Downgraded two levels due to very serious concerns regarding imprecision of effect estimate.
^d Downgraded one level due to serious concerns regarding imprecision of effect estimate.
^e As the heterogeneity between studies seemed to relate to magnitude rather than direction of effect, we did not downgrade for inconsistency.

Background

Oral submucous fibrosis (OSF) is a chronic disease of the oral cavity. It is characterised by the progressive build‐up of constricting bands of collagen in the cheeks and adjacent structures of the mouth, which can cause problems with speech, swallowing, and can severely restrict mouth opening and tongue movement. Systematic reviews estimate a rate of 4.6% to 5.2% of malignant transformation reported in longitudinal studies of OSF, with an annual transformation rate of 0.73% to 0.98% (Iocca 2020; Kujan 2020). The mechanism by which malignant transformation occurs is not well understood.

The prevalence of OSF varies widely between countries, and even between communities and geographical regions within those countries, ranging from 0.1% to 30% (Rao 2020). One epidemiological study in central India noted that there was a prevalence of nearly five times as many males with OSF as females, though incidence was higher in females (Hazarey 2007). In Taiwan, one study demonstrated that there had been an increase in prevalence from 8.3 per 100,000 in 1996 to 16.2 per 100,000 in 2013 (Yang 2018). Worldwide estimates indicate that at least five million people may be affected (Aziz 2008). The condition is more commonly found in people in the Asian subcontinent and the Far East, but, as a result of transmigration patterns, an increasing number of cases are being seen in other countries (Sukumar 2012).

The aetiology of OSF is multifactorial: chewing of the areca nut (from the areca catechu plant), excessive use of chillies and spices, poor nutrition and genetic predisposition have all been suggested as causative factors (Hazarey 2007). Chewing of betel quid (areca nut, lime and tobacco) as well as other areca nut‐containing products (e.g. paan masala and gutka) for mouth freshening and a mild euphoric effect is a relatively common practice in India, Pakistan and Sri Lanka. Other combinations, as well as a variety of packaged products, are available in parts of the Far East, particularly China, Taiwan and Malaysia.

Epidemiological studies have demonstrated an association between chewing of areca nut products and the incidence of OSF (Lee 2011), as well as highlighting the significant contribution of smoking and alcohol consumption to the malignant transformation of OSF (Liu 2015).

The precise mode of action of the various chemical constituents of areca nut on the oral tissues is unclear. It has been suggested that these constituents interfere with the deposition or breakdown of collagen, with the cytokine tumour growth factor‐beta (TGF‐β) being key in mediating the OSF pathway (Arakeri 2017a). Other genetic and autoimmune factors may have a bearing on the pathogenesis of OSF. There are reports that copper from sources such as areca nut and drinking water may contribute to the pathogenesis of OSF (Angadi 2011; Arakeri 2014). Research indicates that the absence of tobacco in preparations of areca/betel quid, as used in Taiwan, does not reduce the likelihood of developing OSF (Yang 2001).

Description of the condition

Symptoms of OSF vary and may include progressive restriction of swallowing and mouth opening, pain, burning sensation exacerbated by spicy food and xerostomia.

Clinical findings in OSF are largely dependent on the severity of the disease, with early stages typically presenting with stomatitis, mucosal ulcers and fibrosis of ruptured vesicles and ulcers. The later stages involve the presence of fibrous bands in the cheeks and lips; depigmented gums and mucosa; rubbery and deformed soft palate and uvula; and a blanched, leathery floor of the mouth (Pindborg 1989).

A number of classification and staging systems for OSF have been used to help determine the type of intervention required (More 2012). Diagnosis of OSF should be based on a thorough history and clinical examination, and confirmed by histopathology of the lesion. Several grading systems have been proposed to classify people with OSF according to disease severity. However, there is no consensus between centres about which grading system is most appropriate.

Description of the intervention

Medical treatment options for OSF include inflammatory modulators; vitamin and mineral supplements; antioxidants; and a range of medicines that promote tissue perfusion or encourage fibrolytic activity, such as intralesional steroids, colchicine, immunised milk, hyaluronidase, human placental extracts, chemotrypsin, pentoxifylline and collagenase. In addition, anti‐TGF‐β agents, copper chelators, turmeric, curcumin and aloe vera have been posited as potential treatments for OSF.

Where there is restriction of mouth opening, surgical interventions have been employed. Release of fibrotic bands has been used in mild cases, whereas masticatory myotomy or coronoidectomy, followed by resurfacing with flaps such as buccal fat pad, nasolabial or radial forearm free flaps, have been used in more extreme cases.

Physical therapy such as mouth opening devices or physiotherapy can also be provided alone or as an adjunct to medical and surgical intervention.

All treatment regimens should include patient education and areca nut habit cessation. If used in the early stages of OSF, such management may result in reduction or alleviation of symptoms (Arakeri 2017b).

There is no reported minimum clinically important difference (MCID) in mouth opening for people with OSF. An absolute change of 10 mm has been suggested as the MCID in a temporomandibular joint dysfunction population (Kaur 2022).

There is no reported MCID in visual analogue scale (VAS) scores for burning sensation experienced by people with OSF. A systematic review of MCIDs in chronic pain conditions has suggested that a relative reduction in VAS scores of 32% may represent an important improvement in symptoms to people experiencing chronic pain (Olsen 2018). In lieu of validated MCID values for people with OSF, we have used this as an estimate of clinically important results.

How the intervention might work

Medical management of OSF, both topical and systemic, may improve symptoms by several possible mechanisms. These include correcting nutritional deficiencies, promoting normal tissue regeneration, fibrinolysis, modulating antifibrotic immune pathways, and promoting blood flow through anticoagulant and vasodilatory activity (Kerr 2011).

Surgical interventions act through mechanical means to improve mouth opening, enabling resumption of speech, normal eating and chewing, as well as facilitating monitoring for early cancerous lesions (Arakeri 2017b).

Physical therapy may influence tissue remodelling through physical movements and localised heat application (Kerr 2011).

Why it is important to do this review

At the time this review was first published in 2008, there were six studies, of which only two were randomised controlled trials (RCTs) that were acceptable for inclusion (Kumar 2007; Rajendran 2006). It was not possible to combine the results in meta‐analyses and these findings were reported narratively. The authors of the original review were unable to draw firm conclusions due to methodological and reporting issues. Since then, numerous new trials have been published. Furthermore, the prevalence of OSF is increasing and the significant impact on the quality of life of individuals is clearer. There are substantial management challenges in treatment of OSF and uncertainty surrounding the most appropriate treatment modalities.

Objectives

To evaluate the benefits and harms of interventions for the management of oral submucous fibrosis.

Methods

Criteria for considering studies for this review

Types of studies

We considered only randomised controlled trials (RCTs) for inclusion in this review. We excluded studies that did not include participants with biopsy‐confirmed OSF or studies that used split‐mouth designs.

Types of participants

People of any age with a confirmed diagnosis, by clinical examination and biopsy, of OSF.

Types of interventions

As there is no established standard of care in OSF, we considered our main findings to be any intervention compared with placebo. The interventions of interest included systemic, intralesional or topical drugs, of any dosage, duration or delivery method; surgical procedures; and others such as physiotherapy, local heat therapy, jaw stretching exercises. We also considered comparisons of these interventions against each other.

Types of outcome measures

Primary outcomes

• Participant‐reported resumption of normal eating, chewing and speech.

• Change or improvement in maximal mouth opening, measured as the interincisal distance.

• Improvement in range of jaw movement using any validated assessment tool.

• Change in severity of oral/mucosal burning pain/sensation using any recognised validated pain scale.

• Adverse effects related to any clinically diagnosed reactions to any of the active interventions (as reported in the included studies)

Secondary outcomes

• Quality of life as assessed by any validated questionnaire, either generic or oral health specific.

• Postoperative discomfort or pain as a result of the intervention: participant‐assessed using any validated pain scale.

• Participant satisfaction assessed by a validated questionnaire.

• Hospital admission: length of stay.

• Direct costs of medication, hospital bed days and any associated inpatient costs for the surgical interventions.

Search methods for identification of studies

Electronic searches

Cochrane Oral Health's Information Specialist conducted systematic searches in the following databases for RCTs. There were no language, publication year or publication status restrictions.

• Cochrane Oral Health's Trials Register (searched 5 September 2022) (Appendix 1)

• Cochrane Central Register of Controlled Trials (CENTRAL; 2022, Issue 8) in the Cochrane Library (Appendix 2)

• MEDLINE Ovid (1946 to 5 September 2022) (Appendix 3)

• Embase Ovid (1980 to 5 September 2022) (Appendix 4)

Dr Xinyu Wu and Dr Rui Lu translated non‐English language papers.

We modelled subject strategies on the search strategy designed for MEDLINE Ovid. Where appropriate, we combined them with subject strategy adaptations of the highly sensitive search strategies designed by Cochrane for identifying RCTs and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions (Lefebvre 2022)).

Searching other resources

Cochrane Oral Health's Information Specialist searched the following trial registries for ongoing studies.

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (clinicaltrials.gov; searched 5 September 2022) (Appendix 5)

• World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch; searched 5 September 2022) (Appendix 6)

We searched the reference lists of relevant articles, and where possible, we contacted the investigators by email to request further trial details and information about any additional trials.

Data collection and analysis

Selection of studies

Three review authors (AJ, TL and BV) independently undertook screening by title and abstract, and obtained full‐text copies of all potentially relevant articles. Three review authors (AJ, TL and BV) then independently screened the full‐text articles for compliance with the inclusion and exclusion criteria, following initial piloting of these criteria to ensure consistency of application. We resolved any discrepancies through consensus, and then through inclusion of fourth and fifth review authors (JT and VRA) where agreement was not reached. Reasons for exclusion are noted in the Characteristics of excluded studies table.

Data extraction and management

At least two review authors (AJ, TL and BV) independently collected study details and outcome data using a predetermined form designed for this purpose. Study details can be found in the Included studies section and Characteristics of included studies table. We included data only if consensus was reached. We discussed any disagreements and, if required, consulted a third or fourth review author. We extracted the following details.

• Study methods: country of conduct, method of allocation, blinding of participants and investigators (if feasible), exclusion of participants after randomisation, numbers lost to follow‐up.

• Participants: sample size, age, sex, inclusion and exclusion criteria.

• Intervention: type of procedure, mode of administration, dose, frequency and duration of usage (for medicinal interventions).

• Control: type, dose and frequency of any comparison or placebo.

• Outcomes: primary and secondary outcomes.

• Adverse effects/events.

If stated, we recorded the sources of funding for any of the included studies.

Assessment of risk of bias in included studies

At least two review authors (AJ, BV and TL) independently assessed the risk of bias of included trials using the Cochrane RoB 1 tool (Higgins 2011). We considered the following factors: random sequence generation, allocation concealment, blinding of participants and investigators, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. We categorised each domain as low, high or unclear risk of bias. We compared the judgements of different review authors and resolved any inconsistencies through discussion. Where we judged the risk of bias to be unclear, we contacted trial authors for clarification.

We made an overall judgement of risk of bias for each study. We assessed studies to have an overall low risk of bias when any plausible bias across all seven domains was unlikely to have altered the results. We assessed studies to have an overall unclear risk of bias when any plausible bias across one or more of the key domains raised some doubt that it may have altered the results. We assessed studies to have an overall high risk of bias when we believed any plausible bias across one or more of the key domains may seriously have altered the results reported in that study.

Measures of treatment effect

We used Review Manager 5 to perform the analyses (Review Manager 2020).

For continuous data, we calculated mean differences (MD) and standard deviations for each group to express the estimate of effect as MD with 95% confidence interval (CI). When studies reported continuous outcomes on different scales, we used standardised mean difference (SMD) to pool these data in meta‐analysis. We calculated odds ratios and their 95% CIs for dichotomous data where possible.

Where there was insufficient information to enable effect measures to be calculated, we provided a narrative report of the summary measures.

Unit of analysis issues

The unit of analysis was the individual participant.

We grouped time points for outcomes into less than three months, three to six months, and greater than six months, based on the most frequently reported follow‐up intervals used in the studies evaluated. Where there was a choice of time point within these groupings for a primary outcome, we selected the longest follow‐up period available. Where there were multiple‐arm studies, we ensured participants were not double counted in meta‐analyses.

Dealing with missing data

We attempted to contact study authors to retrieve any partial or missing data. Where complete data were not available, we planned to impute data when it was deemed sufficiently low risk to do so. No imputation of data was carried out.

Assessment of heterogeneity

We assessed clinical heterogeneity by examining the characteristics of the studies: the similarity between the types of participants, the interventions and the outcomes as specified in the Criteria for considering studies for this review.

We assessed statistical heterogeneity by calculation of the Q statistic with P value set at less than 0.10. This was quantified by the calculation of the I² statistic for heterogeneity. We judged values above 75% to represent considerable heterogeneity and values from 50% to 74% as substantial, based on guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

We pooled results of trials to provide estimates of the effectiveness of the interventions only if we judged that the included studies were sufficiently homogeneous with respect to intervention received and study participant characteristics, using the random‐effects model of meta‐analysis.

Assessment of reporting biases

Had we found a sufficient number of studies within each intervention group to enable reliable assessment by funnel plots or Egger's test, we would have used these to consider any possible reporting bias.

Data synthesis

We followed Cochrane statistical guidelines for data synthesis. Review authors (AJ, JT and VRA) analysed the data using Review Manager 5 and reported them according to Cochrane criteria (Higgins 2011; Review Manager 2020).

We performed meta‐analysis using the random‐effects model. Forest plots displayed means and 95% CIs. Where it was not possible to pool data or where data reporting did not allow for quantitative analysis, we provided a narrative description of the findings.

Subgroup analysis and investigation of heterogeneity

We subgrouped studies according to the putative primary mechanism of action of the intervention. Because interventions between subgroups differed substantially in mode of delivery and mechanism of action, we considered it inappropriate to undertake comparisons between subgroups.

We had planned to undertake a subgroup analysis based on staging of OSF. We did not find sufficient data to undertake this.

Sensitivity analysis

As all but one study was at high or unclear risk of bias, we did not undertake a sensitivity analysis.

Summary of findings and assessment of the certainty of the evidence

We prepared a summary of review table, which presented each comparison and the primary outcomes. As there is no established standard of care in OSF, we considered 'any intervention versus placebo' to be our main comparison of interest (Table 1).

Two review authors (AJ and JT) assessed the certainty of evidence using GRADEpro GDT software as high, moderate, low or very low (GRADEpro GDT). Factors that were assessed and contributed to this evaluation using the GRADE methodology were: risk of bias, inconsistency between studies, indirectness, imprecision and likely publication bias.

Results

Description of studies

See Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; and Characteristics of ongoing studies tables.

Results of the search

Up to September 2022, the search identified 1108 references for this version of the review from the electronic searches. After removing duplicates, 605 references remained. We considered 121 articles in full text and excluded 79 of these, with reasons. We included 28 new studies (29 references). We included two studies from a previous version of this review. Therefore, we included 30 RCTs (31 records) in this review (Figure 1). Three studies are awaiting classification, and there are eight ongoing studies.

1.

Included studies

For a summary of each included study, see the Characteristics of included studies table.

Characteristics of the trials

Design

Twenty‐two RCTs used a two‐arm parallel‐group design (Ara 2018; Arakeri 2020; Dani 2018; Kania 2022; Mehrotra 2011; Mulk 2013; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b; Patil 2017; Patil 2019; Piyush 2018; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Singh 2010; Singh 2016; Sudarshan 2012; Thakur 2015; Vadepally 2019; Yadav 2014), seven used a three‐arm parallel‐group design (Alora Veedu 2015; Bhadage 2013; Cox 2009; Goel 2015; Jiang 2013; Jirge 2008; Kumar 2007), and one used a four‐arm parallel‐group design (Wu 2010).

Setting

Twenty‐seven studies were conducted in India (Alora Veedu 2015; Ara 2018; Arakeri 2020; Bhadage 2013; Dani 2018; Goel 2015; Jirge 2008; Kania 2022; Kumar 2007; Mehrotra 2011; Mulk 2013; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b; Patil 2017; Patil 2019; Piyush 2018; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Singh 2010; Singh 2016; Sudarshan 2012; Thakur 2015; Vadepally 2019; Yadav 2014), two in China (Jiang 2013; Wu 2010), and one in Nepal (Cox 2009). All studies were single centre and conducted in university clinics or hospitals.

Funding

Seven studies reported that they received no funding (Ara 2018; Dani 2018; Goel 2015; Jirge 2008; Patil 2014a; Patil 2015a; Singh 2016), one study received institutional funding (Jiang 2013), and two studies received funding from state medical boards (Alora Veedu 2015; Wu 2010). Twenty studies did not make funding declarations (Arakeri 2020; Bhadage 2013; Cox 2009; Kania 2022; Kumar 2007; Mehrotra 2011; Mulk 2013; Patil 2014b; Patil 2015b; Patil 2017; Patil 2019; Piyush 2018; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Singh 2010; Sudarshan 2012; Thakur 2015; Vadepally 2019; Yadav 2014).

Participants

The total number of participants included in the trials was 2176, with the number per study ranging from eight to 400. Studies did not consistently report baseline characteristics of participants such as age, sex or disease stage.

Interventions

Table 2 shows a summary of the interventions used in the studies.

1. Characteristics of the interventions.

Intervention	Delivery method	Details of intervention	Study
Steroids (alone or in combination)
Betamethasone	Local	Intralesional injection of betamethasone 4 mg/mL diluted in 1 mL of 2% xylocaine bilaterally biweekly for 6 months + habit cessation advice and oral prophylaxis	Goel 2015
Dexamethasone	Local	Submucosal injections of dexamethasone 2 mL (4 mg/mL) with 1 mL of 2% lignocaine with adrenaline (1:80,000), bilateral buccal mucosa, biweekly for 5 weeks with cheek massage and mouth opening physiotherapy 5 times daily and passive tongue spatula placement between teeth	Alora Veedu 2015
Combination hyaluronidase + dexamethasone	Local	Submucosal injections of a combination of dexamethasone 1 mL (4 mg/mL) and hyaluronidase (750 IU) diluted with 1 mL of 2% lignocaine with adrenaline (1:80,000) and 1 mL of water for injection. Bilateral buccal mucosa, biweekly for 5 weeks with cheek massage and mouth opening physiotherapy 5 times daily and passive tongue spatula placement between teeth	Alora Veedu 2015
Hydrocortisone and hyaluronidase	Local	Injection of a combination of hydrocortisone acetate (1.5 mL of 25 mg/mL) and hyaluronidase (1500 IU) submucosally in retromolar trigone, soft palate and cheek bilaterally (half dose on each side). Performed every 15 days for 22 weeks	Singh 2010
Triamcinolone and hyaluronidase	Local	Injection of a combination of triamcinolone acetonide (10 mg/mL) and hyaluronidase (1500 IU) submucosally in retromolar trigone, soft palate and cheek bilaterally (half dose on each side). Performed every 15 days for 22 weeks	Singh 2010
Dexamethasone + hyaluronidase + 2% lidocaine	Local	Intralesional injection 2 mL of dexamethasone with 1500 IU hyaluronidase mixed with 1 mL of 2% lidocaine with 1:80,000 adrenaline twice‐weekly for 6 weeks with mouth opening physiotherapy, cheek blowing exercises and habit cessation advice	Bhadage 2013
Hydrocortisone + hyaluronidase	Local	Intralesional injection of hyaluronidase 1500 U and hydrocortisone 100 mg in sterile water, biweekly for 4 weeks with habit cessation advice and correction of any underlying anaemia with mebendazole 100 mg twice daily for 3 days and ferrous sulphate 200 mg and folic acid 5 mg for 4 months	Cox 2009
Oral lycopene + betamethasone	Local + Systemic	Lycopene 16 mg orally daily + 2 × 4 mg intralesional injection of betamethasone for 2 months with habit cessation advice	Kumar 2007
Triamcinolone	Local	Intralesional injection of triamcinolone 2 mg after 5 minutes of local anaesthetic cream, once per week for 20 weeks with habit cessation advice, and mouth stretching exercises 3 times per day for 15 minutes	Jiang 2013
Salvianolic acid + triamcinolone	Local	Intralesional injection of both triamcinolone 2 mg and salvianolic acid B 4 mg after 5 minutes of local anaesthetic cream application, at weekly intervals for 20 weeks with habit cessation advice, and mouth stretching exercises 3 times per day for 15 minutes	Jiang 2013
Dexamethasone + hyaluronidase	Local	Intralesional injection of dexamethasone 4 mg/mL, hyaluronidase 1500 IU and 0.5 mL of 2% lidocaine with adrenaline administered weekly at multiple sites for 3 months	Yadav 2014
Prednisolone + salvia miltorrhiza	Local	Weekly intralesional injections of 2 mL prednisolone acetate (25 g/L) and 1 mL lidocaine (20 g/L) and 2 mL of compound salvia miltorrhiza (1.5 g/mL) and 1 mL lidocaine (20 g/L) for 3 months	Wu 2010
Prednisolone	Local	Weekly intralesional injection of prednisolone acetate (25 g/L) and 1 mL lidocaine (20 g/L) for 3 months	Wu 2010
Hyaluronidase + hydrocortisone	Local	Injection of hyaluronidase 1500 U and hydrocortisone 100 mg in sterile water biweekly for 4 weeks with habit cessation advice, diet advice and correction of underlying anaemia with mebendazole 100 mg twice daily for 3 days and ferrous sulphate 200 mg and folic acid 5 mg for 4 months	Cox 2009
Surgery (alone or in combination)
Surgery – extended nasolabial flaps	Local	Bilateral incision and release of buccal fibrous bands, coronoidectomy, reconstruction with extended nasolabial flaps, mouth opening physiotherapy for 6 months postoperatively	Patil 2017
Surgery – nasolabial flap	Local	Excision of fibrous bands, mouth opening using Heister's mouth gag (35–40 mm), coronoidectomy and reconstruction with nasolabial flap. Extraction of third molars, intensive physiotherapy using Heister's mouth gag 5–6 times daily for 20–30 minutes from day 3 postoperatively onwards	Kania 2022
Surgery – abdominal dermal fat graft	Local	Excision of fibrous bands, mouth opening using Heister's mouth gag (35–40 mm), coronoidectomy, and reconstruction with abdominal dermal fat graft. Extraction of third molars, intensive physiotherapy using Heister's mouth gag 5–6 times daily for 20–30 minutes from day 3 postoperatively onwards	Kania 2022
Surgery – buccal fat pad graft	Local	Bilateral incision and release of buccal fibrous bands, coronoidectomy, reconstruction with buccal fat pad graft, mouth opening physiotherapy for 6 months postoperatively	Patil 2017
Surgery + placental extract	Local	Bilateral buccal mucosal fibrotomy and coronoidectomy repaired with buccal fat pad. Placental extract applied topically 4 hourly for 4 weeks with physiotherapy and oral hygiene instruction	Thakur 2015
Surgery	Local	Bilateral buccal mucosal fibrotomy and coronoidectomy repaired with buccal fat pad with physiotherapy and oral hygiene instruction	Thakur 2015
Surgery + Heister jaw opener	Local	Mouth opening physiotherapy using Heister jaw opener instrument between teeth 5–6 times daily for 6 months followed by another 12 months at reduced frequency alongside fibrous band incision with buccal fat pad repair, bilateral coronoidectomy, extraction of third molars, oral supplements of vitamin A 50,000 IU, vitamin B complex 200 mg, vitamin C 500 mg and beta‐carotene, along with topical application of triamcinolone acetonide 0.1% applied over the surgical area for ≥ 6 months postoperatively	Vadepally 2019
Surgery + wooden tongue depressors	Local	Mouth opening physiotherapy using placement of wooden tongue depressors between teeth 5–6 times daily for 6 months, followed by another 12 months at reduced frequency. Fibrous band incision with buccal fat pad flap repair, bilateral coronoidectomy, and extraction of third molars. Oral supplements of vitamin A 50,000 IU, vitamin B complex 200 mg, vitamin C 500 mg and beta‐carotene, along with topical application of triamcinolone acetonide 0.1% applied over the surgical area for ≥ 6 months postoperatively	Vadepally 2019
Physiotherapy (alone or in combination)
Physiotherapy	Local	Tongue spatulas placed passively between anterior teeth holding jaws open for 1 minute 5 times per session 5 times per day for 4 months with the addition of an additional spatula every 5th day with habit cessation advice, diet advice and correction of any underlying anaemia with mebendazole 100 mg twice daily for 3 days and ferrous sulphate 200 mg and folic acid 5 mg for 4 months	Cox 2009
Physiotherapy + ultrasound therapy	Local	Ultrasound therapy for 7 minutes over 1 week (intervals unknown) + jaw opening exercises – maximal opening for 5 seconds, lateral deviations and protrusion of mandible repeated 10 times	Dani 2018
Physiotherapy	Local	Jaw opening exercises – maintain maximal mouth opening for 5 seconds, lateral deviations and protrusion of mandible, repeated 10 times for 1 week	Dani 2018
Pentoxifylline
Pentoxifylline	Systemic	Oral pentoxifylline 400 mg twice daily for 30 days then 3 times daily for 6 months + habit cessation advice	Mehrotra 2011
Pentoxifylline	Systemic	Oral pentoxifylline 2 × 400 mg tablets daily for 30 days then 3 × 400 mg tablets daily for 6 months with 1 multivitamin tablet at night, local heat therapy and mouth opening exercises	Rajendran 2006
Pentoxifylline	Systemic	2 × pentoxifylline 400 mg tablets twice daily for 3 months with habit cessation advice	Patil 2014a
Pentoxifylline	Systemic	2 × pentoxifylline tablets daily for 15 days, followed by 3 tablets daily for 3.5 months	Prabhu 2015
Pentoxifylline	Systemic	Pentoxifylline 400 mg orally twice daily for 3 months with daily mouth opening exercises	Mulk 2013
Aloe vera
Aloe vera	Topical	Aloe vera 5 mg gel applied topically to buccal mucosa bilaterally 3 times per day for 3 months with habit cessation advice	Sudarshan 2012
Aloe vera	Topical	Aloe vera 5 mg gel topically 3 times daily for 3 months	Patil 2015b
Aloe vera	Topical	Aloe vera gel applied topically to lesions 3 times daily for 3 months with habit cessation advice	Patil 2014b
Lycopene
Lycopene	Systemic	Lycopene 2 mg capsules twice daily for 6 months with habit cessation advice and oral prophylaxis	Goel 2015
Lycopene	Systemic	Lycopene 16 mg orally daily for 2 months with habit cessation advice	Kumar 2007
Lycopene	Systemic	1 × lycopene 8 mg capsule twice daily for 6 months	Piyush 2018
Mixed antioxidants
Mixed antioxidants	Systemic	Antioxidant capsules (beta‐carotene 5 mg; mixed carotenoids 5 mg; vitamin E 25 mg; chromium 200 μg; zinc 7.5 mg; manganese 1.5 mg; copper 1 mg; selenium 150 μg; vitamin C 100 mg) orally twice daily for 3 months with habit cessation advice	Sudarshan 2012
Mixed antioxidants	Systemic	1 tablet of Turmix (combination curcumin and piperine 300 mg) twice daily for 6 months	Piyush 2018
Mixed antioxidants	Systemic	2 tablets of Turmix (Curcumin longa 300 mg, piperine 5 mg) orally for 3 months	Yadav 2014
Oxitard
Oxitard	Systemic	2 Oxitard capsules twice daily for 3 months	Patil 2015a; Patil 2019
Oxitard	Systemic	2 Oxitard capsules twice daily for 3 months with habit cessation advice	Patil 2014b
Spirulina
Spirulina	Systemic	Spirulina 250 mg orally twice daily for 3 months	Patil 2015b
Spirulina	Systemic	Spirulina 500 mg orally twice daily for 3 months with daily mouth opening exercises	Mulk 2013
Spirulina	Systemic	Spirulina 500 mg twice daily for 3 months with habit cessation advice	Patil 2019
Other
Curcumin	Systemic	2 curcumin 500 mg capsules daily + habit cessation for 6 months	Ara 2018
Isoxsuprine	Systemic	Isoxsuprine 10 mg orally 4 times per day for 6 weeks, mouth opening physiotherapy and cheek blowing exercises, habit cessation advice	Bhadage 2013
Xantitol nicotinate	Local	Intralesional injection of xantitol nicotinate biweekly for 4 months with mouth opening physiotherapy	Singh 2016
Salvianolic acid	Local	Intralesional injection of salvianolic acid B 4 mg after 5 minutes of local anaesthetic cream application at weekly intervals for 20 weeks with habit cessation advice, and mouth stretching exercises 3 times per day for 15 minutes	Jiang 2013
Salvianolic acid	Systemic	1 capsule salvianolic acid antioxidant twice daily for 6 weeks with habit cessation advice	Jirge 2008
Levisamole	Systemic	Levisamole 50 mg tablet orally 3 times daily for 3 consecutive days per week for 3 alternate weeks with habit cessation advice	Jirge 2008
Levisamole + salvianolic acid	Systemic	Levisamole 50 mg tablet orally 3 times daily for 3 consecutive days per week for 3 alternate weeks with 1 capsule salvianolic acid twice daily for 6 weeks + habit cessation advice	Jirge 2008

Fourteen trials tested preparations with proposed antioxidant properties administered orally. These were: Oxitard (Patil 2014b; Patil 2015a; Patil 2019), lycopene (Arakeri 2020; Goel 2015; Kumar 2007; Piyush 2018), spirulina (Mulk 2013; Patil 2015b; Patil 2019), 'Turmix' capsules (Piyush 2018; Yadav 2014), salvianolic acid (Jiang 2013), curcumin (Ara 2018; Rajbhoj 2021), and 'antioxidant capsules' (Lekar Pharma Ltd) (Sudarshan 2012).

Four trials used a single corticosteroid injected locally (betamethasone, dexamethasone, triamcinolone, prednisolone and hydrocortisone) (Alora Veedu 2015; Goel 2015; Jiang 2013; Wu 2010). Five trials used a steroid in combination with hyaluronidase injected locally (Alora Veedu 2015; Bhadage 2013; Cox 2009; Singh 2010; Yadav 2014). One trial used a steroid injection combined with salvianolic acid (Jiang 2013). One trial compared a corticosteroid injected locally with oral lycopene (Kumar 2007).

Five trials employed systemic pentoxifylline (an anti‐inflammatory and innate immunomodulatory drug) (Mehrotra 2011; Mulk 2013; Patil 2014a; Prabhu 2015; Rajendran 2006).

Two trials employed vasodilators injected locally; these were isoxsuprine (Bhadage 2013) and xanthinol nicotinate (Singh 2016).

One trial compared levamisole (a drug with proposed immunomodulatory properties) with levamisole combined with an antioxidant (Jirge 2008).

Two trials compared two forms of surgery: one trial compared buccal fat pad graft versus extended nasolabial flaps (Patil 2017), and one trial compared abdominal dermal fat graft versus nasolabial flap (Kania 2022). One trial compared surgery versus surgery plus a topical application of placental extract (Thakur 2015). One trial compared two forms of physiotherapy after surgery (Heister jaw opener versus wooden tongue depressors) (Vadepally 2019).

Two trials compared different forms of physiotherapy (Cox 2009; Dani 2018).

Three trials included topical applications of aloe vera gel as a comparator (Patil 2014b; Patil 2015b; Sudarshan 2012).

Thirteen trials compared interventions with a placebo or non‐active control: seven antioxidant studies (Arakeri 2020 (lycopene); Ara 2018 (curcumin); Kumar 2007 (lycopene); Mulk 2013 (spirulina); Patil 2015a (Oxitard); Piyush 2018 (Turmix – curcumin and piperine 300 mg); Yadav 2014 (Turmix – Curcumin longa 300 mg, piperine 5 mg)); two pentoxifylline studies (Patil 2014a; Prabhu 2015); two vasodilator studies (Bhadage 2013; Singh 2016); one physiotherapy study (Cox 2009); and one aloe vera study (Patil 2014b).

Treatment courses varied from one week to 18 months.

We evaluated the following comparisons by meta‐analysis.

• Any intervention versus placebo

• Different types of medications versus one another

• Different surgical techniques versus one another

• Surgery alone versus surgery plus adjunctive treatments

• Physiotherapy alone versus physiotherapy plus adjunctive treatments

• Physiotherapy versus medications

• Surgery combined with different physiotherapy techniques

Outcomes

No studies reported our primary outcome of interest of participant‐reported resumption of normal eating, chewing and speech. Four trials reported change in 'difficulty in speech', 'difficulty in swallowing', 'difficulty in mouth opening' and 'intolerance to spicy food' (Ara 2018; Patil 2014a; Patil 2014b; Patil 2015a). These were presented as frequencies of participants reporting these symptoms before and after interventions. As these are important participant‐reported outcomes, Table 3 provides a summary. However, interpretation of these findings was difficult. Interincisal distance (change scores, measured in millimetres) and severity of burning sensation (measured using a VAS from 0 mm to 100 mm) were the outcome measures most frequently reported by trials included in this review. Several studies reported pain as distinct from burning sensation, although it was not clear how these were differentiated. The outcomes were measured between one week and 18 months. No trials reported our secondary outcomes.

2. Participant‐reported outcome measures.

Study	Population	Intervention	Comparison	Follow‐up	Outcome (s)	Intervention – outcome frequency (%) (baseline – endpoint)	Control – outcome frequency (%) (baseline – endpoint)	Comment
Ara 2018	100 (50 per group)	Curcumin	Placebo	6 months	Difficulty in mouth opening	88	0	Difference between groups (P < 0.001)
					Difficulty in swallowing	8	0	Difference between groups (P < 0.001)
					Intolerance to spicy food	42	0	Difference between groups (P < 0.001)
Patil 2014b	120	Oxitard capsules	Topical aloe vera gel	5 months	Difficulty in swallowing	83	55	Difference between groups (P < 0.001)
					Difficulty in speech	76	59	Difference between groups (P < 0.001)
Patil 2014a	106	Pentoxifylline 400 mg twice daily	Multivitamin (placebo)	5 months	Difficulty in swallowing	62	43	Intergroup differences not assessed
					Difficulty in speech	58	45	Intergroup differences not assessed
Patil 2015a	120	2 Oxitard capsules twice daily	Placebo	3 months	Difficulty in swallowing	83	62	Difference between groups (P < 0.001)
					Difficulty in speech	76	64	Difference between groups (P < 0.001)

Maximal mouth opening (interincisal distance (mm))

Twenty‐nine studies used interincisal distance to measure a change or improvement in maximal mouth opening. The units of measurement were either millimetres or centimetres. We converted results to millimetres for quantitative analysis.

Burning sensation

Fourteen studies used an assessment of burning sensation measured using a validated scale (Alora Veedu 2015; Ara 2018; Arakeri 2020; Bhadage 2013; Jiang 2013; Jirge 2008; Mulk 2013; Piyush 2018; Prabhu 2015; Rajbhoj 2021; Singh 2016; Sudarshan 2012; Vadepally 2019; Yadav 2014). All studies used a VAS, from either 0 cm to 10 cm or 0 mm to 100 mm. We converted results to millimetres for quantitative analysis.

Eight studies reported burning sensation using only non‐validated scales, subjective scores, as a binary measure of presence/absence or were unclear about how this outcome was measured (Kumar 2007; Mehrotra 2011; Patil 2014a; Patil 2015a; Patil 2015b; Patil 2019; Rajendran 2006; Singh 2010).

Eight studies did not measure burning sensation (Cox 2009; Dani 2018; Goel 2015; Kania 2022; Patil 2014b; Patil 2017; Thakur 2015; Wu 2010).

Pain

Six studies reported presence of pain as a binary outcome (Alora Veedu 2015; Ara 2018; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b). Table 4 shows a summary of these findings.

3. Pain.

Study	Population	Intervention	Comparator(s)	Follow‐up	Outcomes (s)	Intervention A – outcome mean difference (SD)	Intervention B – outcome mean (SD)	Comparison – outcome mean difference (SD)	Comment
Alora Veedu 2015	45	Hyaluronidase submucosal injection	Dexamethasone submucosal injection Hyaluronidase + dexamethasone submucosal injections	6 months	Pain while opening mouth (VAS)	1.00 (1.69)a	0.33 (0.72)a	1.93 (2.15)a	Difference between groups (P = 0.035)
Ara 2018	100	2 × 500 mg curcumin capsules once daily	Placebo	6 months	Pain (VAS)	3.16 (4.23)		0.00	Unclear how pain was differentiated from burning sensation in this study
Study ID	Population	Intervention	Comparator(s)	Follow‐up	Outcomes (s)	Intervention A – outcome frequency (%) (baseline – endpoint)	—	Comparison – outcome frequency (%) (baseline –endpoint)	Comment
Patil 2014a	106	Pentoxifylline 400 mg twice daily	Multivitamin (placebo)	5 months	Pain associated with lesion	75	—	58	—
Patil 2014b	120	2 Oxitard capsules twice daily	Topical aloe vera gel 5 mg 3 times daily	5 months	Pain associated with lesion	79	—	48	—
Patil 2015a	120	2 Oxitard capsules twice daily	Placebo	3 months	Pain associated with lesion	79	—	62	—
Patil 2015b	42	Spirulina 250 mg oral twice daily	Topical aloe vera gel 5 mg 3 times daily	5 months	Pain associated with lesion	30	—	57	Intergroup differences not measured

^a change from baseline reported
SD: standard deviation; VAS: visual analogue scale.

Adverse effects

Fifteen studies made statements regarding adverse effects. Of these, six did not encounter adverse effects of treatment (Bhadage 2013; Jiang 2013; Patil 2014b; Patil 2019; Singh 2010; Yadav 2014), while nine reported adverse effects in treatment groups (Kania 2022; Kumar 2007; Mulk 2013; Patil 2015b; Patil 2017; Piyush 2018; Singh 2016; Sudarshan 2012; Vadepally 2019). The 15 remaining studies either did not consider adverse effects at all or did not report data in a usable form (Alora Veedu 2015; Ara 2018; Arakeri 2020; Cox 2009; Dani 2018; Goel 2015; Jirge 2008; Mehrotra 2011; Patil 2014a; Patil 2015a; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Thakur 2015; Wu 2010).

Other outcomes

The studies did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Outcomes not relevant for this review

Trials reported several other outcomes that we did not deem reliable or relevant to this review. See Appendix 7.

Studies without usable data

We excluded seven studies from meta‐analysis as data were mis‐reported (Mehrotra 2011) or not reported in a usable format (Alora Veedu 2015; Goel 2015; Patil 2015b; Rajendran 2006; Singh 2010; Sudarshan 2012).

Excluded studies

Although many of the excluded studies could have been excluded for more than one reason, we recorded only the main reason for exclusion in the Characteristics of excluded studies table.

We excluded 39 studies that did not use a histological diagnosis for OSF (Ambereen 2021; Anuradha 2017; Baptist 2016; Biswas 2015; Chandrashekar 2021; Chen 2021; CTRI/2021/09/036499; Daga 2017; Dasukil 2022; Datarkar 2020; Deepak 2021; Ganguly 2020; Gupta 2022; Hazarey 2015; Jiang 2015; Kanjani 2019; Kisave 2020; Lanjekar 2020; Memon 2022; Mostafa 2021; Nihadha 2022; Pandey 2020; Patil 2016a; Patil 2016b; Pipalia 2016; Rai 2019; Raizada 2017; Raizada 2022; Rizvi 2019; Saalim 2020; Sadaksharam 2017; Saran 2018; Selvam 2013; Shams 2022; Shandilya 2021; Singh 2016a; Singh 2018; Srivastava 2021; Tp 2019); 36 studies as they were not, or were unlikely to be, RCTs (Agarwal 2011; Agrawal 2018; Angadi 2011; Ara 2016; Awan 2014; Aziz 2009; Bande 2013; Bande 2016; Bohra 2021; Borle 2009; Chole 2012; Gupta 1988; Idrees 2016; Jiang 2009; Johny 2019; Kakar 1985; Kamath 2014; Karemore 2012; Kerr 2011; Lambade 2015; Liu 2017; Lu 2019; Mehrotra 2009; Pardeshi 2015; Pentapati 2016; Pentapati 2017; Pipalia 2017; Rai 2014; Sharma 2012; Shetty 2013; Singh 2015; Singh 2016b; Tai 2001; Vibha 2019; Warnakulasuriya 2016; Wollina 2015); two studies that did not measure outcomes relevant to this review (Krishnamoorthy 2013; Mathai 2017); one study that did not have the full text available (Jain 2000); and one study that did not distinguish participants with OSF from those with oral cancer (Li 2018).

Studies awaiting classification

Three studies are awaiting classification (Alam 2013; CTRI/2022/03/041108; Wu 2015).

Ongoing studies

We identified eight ongoing trials, which may be included in future versions of this review (CTRI201202002437; CTRI201305003609; CTRI201504005696; CTRI201701007732; CTRI201709009666; CTRI201709009674; CTRI201804013147; CTRI201804013285).

Risk of bias in included studies

We present details of the assessment of the risk of bias for each included study in the Characteristics of included studies table, Figure 2, and Figure 3.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

3.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Overall risk of bias

One study was at low overall risk of bias (Piyush 2018), five were at unclear overall risk of bias (Alora Veedu 2015; Ara 2018; Arakeri 2020; Patil 2015a; Singh 2016), and the remaining 24 studies were at high overall risk of bias due to inadequate randomisation, allocation concealment, blinding, attrition bias or selective reporting (Bhadage 2013; Cox 2009; Dani 2018; Goel 2015; Jiang 2013; Jirge 2008; Kania 2022; Kumar 2007; Mehrotra 2011; Mulk 2013; Patil 2014a; Patil 2014b; Patil 2015b; Patil 2017; Patil 2019; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Singh 2010; Sudarshan 2012; Thakur 2015; Vadepally 2019; Wu 2010; Yadav 2014).

Allocation

We considered the method of randomisation adequate in both of its components (sequence generation and allocation concealment) in three trials (Dani 2018; Piyush 2018; Vadepally 2019). In two trials, sequence generation was adequate but allocation concealment was at high risk of bias (Kania 2022; Singh 2010). In six trials, sequence generation was adequate but allocation concealment unclear (Alora Veedu 2015; Ara 2018; Mulk 2013; Rajbhoj 2021; Thakur 2015; Yadav 2014). Sequence generation and allocation concealment were both at high risk of bias for one trial (Cox 2009). The remaining 18 trials had unclear sequence generation and allocation concealment (Arakeri 2020; Bhadage 2013; Goel 2015; Jiang 2013; Jirge 2008; Kumar 2007; Mehrotra 2011; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b; Patil 2017; Patil 2019; Prabhu 2015; Rajendran 2006; Singh 2016; Sudarshan 2012; Wu 2010).

Blinding

Performance bias

We judged eight trials at low risk of performance bias, as both participants and personnel were blinded (Alora Veedu 2015; Ara 2018; Arakeri 2020; Mehrotra 2011; Patil 2015a; Piyush 2018; Singh 2010; Singh 2016).

We judged 20 trials at high risk of performance bias (Bhadage 2013; Cox 2009; Dani 2018; Goel 2015; Jiang 2013; Kania 2022; Kumar 2007; Mulk 2013; Patil 2014b; Patil 2015b; Patil 2017; Patil 2019; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Sudarshan 2012; Thakur 2015; Vadepally 2019; Wu 2010; Yadav 2014), and two studies as unclear (Jirge 2008; Patil 2014a).

Detection bias

Mouth opening

All included studies measured mouth opening and all but one reported interincisal distance in millimetres as a measure of mouth opening. We judged studies using this objective measure to be at low risk of detection bias as it was unlikely to be influenced by lack of blinding.

Burning sensation

Outcome assessors and participants were blinded when assessing burning sensation in nine studies, resulting in a low risk of bias judgement (Alora Veedu 2015; Ara 2018; Arakeri 2020; Mehrotra 2011; Patil 2014a; Patil 2015a; Piyush 2018; Singh 2010; Singh 2016). Assessors were not blinded in 13 studies resulting in a high risk of bias judgement (Bhadage 2013; Jiang 2013; Jirge 2008; Kumar 2007; Mulk 2013; Patil 2015b; Patil 2019; Prabhu 2015; Rajbhoj 2021; Rajendran 2006; Sudarshan 2012; Vadepally 2019; Yadav 2014). Eight studies did not assess burning sensation (Cox 2009; Dani 2018; Goel 2015; Kania 2022; Patil 2014b; Patil 2017; Thakur 2015; Wu 2010).

Incomplete outcome data

We judged 19 trials at low risk of attrition bias since all enrolled participants completed the study, or the number of participants lost was unlikely to have had a significant impact on the intervention effect estimate (Ara 2018; Arakeri 2020; Dani 2018; Jiang 2013; Kania 2022; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b; Patil 2017; Patil 2019; Piyush 2018; Prabhu 2015; Rajbhoj 2021; Singh 2010; Singh 2016; Sudarshan 2012; Thakur 2015; Wu 2010). No intention‐to‐treat analyses were performed.

We judged three trials at high risk of attrition bias where the rate of dropouts was higher than 20% (Cox 2009; Kumar 2007; Mulk 2013). A further trial was at high risk of bias as the attrition was very imbalanced between groups (Bhadage 2013).

Seven trials did not report the number of participants completing treatment and so were at unclear risk of attrition bias (Alora Veedu 2015; Goel 2015; Jirge 2008; Mehrotra 2011; Rajendran 2006; Vadepally 2019; Yadav 2014).

Selective reporting

We judged 21 trials at low risk of reporting bias since all planned outcomes were reported (Alora Veedu 2015; Ara 2018; Arakeri 2020; Bhadage 2013; Cox 2009; Dani 2018; Jiang 2013; Jirge 2008; Kania 2022; Mulk 2013; Patil 2015a; Patil 2017; Patil 2019; Piyush 2018; Rajbhoj 2021; Singh 2010; Singh 2016; Sudarshan 2012; Thakur 2015; Wu 2010; Yadav 2014).

Nine studies reported one or more outcomes of interest incompletely or in a way that did not allow adequate interpretation and were at high risk of bias (Goel 2015; Kumar 2007; Mehrotra 2011; Patil 2014a; Patil 2014b; Patil 2015b; Prabhu 2015; Rajendran 2006; Vadepally 2019).

Other potential sources of bias

One study did not report any relevant baseline group characteristics and was judged at unclear risk of other bias (Jirge 2008). The remaining studies were at low risk of other bias.

Effects of interventions

See: Table 1

Any intervention versus placebo or no active treatment

We considered trials comparing any intervention to placebo or no active treatment (not including habit cessation or jaw exercises if they were part of standard care) to be our main results because there is no established standard of care for OSF.

Participant‐reported resumption of normal eating, chewing and speech

No studies reported participant‐reported resumption of normal eating, chewing and speech at any time point. Four studies did report a proxy measure – difficulty in speech and swallowing (Ara 2018; Patil 2014a; Patil 2014b; Patil 2015a). The four studies reported these as frequency of presence or absence of 'difficulty'. This method of reporting subjective outcomes is not validated and cannot differentiate between degrees of improvement that might represent significant functional gains to study participants; however, for completeness, we present the results.

A summary of these studies is reported in Table 3.

Maximal mouth opening (interincisal distance (mm))

Of the placebo‐controlled trials, we identified six subgroups based on the putative mechanism of action of the interventions.

Less than three months' follow‐up

Four study subgroups reported outcomes at follow‐up of less than three months (Analysis 1.1).

1.1. Analysis.

Comparison 1: Any intervention vs placebo (< 3 months' follow‐up), Outcome 1: Interincisal distance (mm)

Two trials with 520 participants compared antioxidants (Oxitard and lycopene) with placebo (Arakeri 2020; Patil 2015a). Antioxidants may increase interincisal distance slightly (MD 3.11 mm, 95% CI 0.46 to 5.77; low‐certainty evidence). There was considerable heterogeneity between these studies (I² = 96%). We downgraded the certainty of the evidence due to high risk of bias (one level) and inconsistency of findings (one level).

One trial with 106 participants compared pentoxifylline versus placebo (Patil 2014a). Pentoxifylline may increase interincisal distance slightly (MD 1.80 mm, 95% CI 1.02 to 2.58; low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

Two trials with 85 participants compared vasodilators (isoxsuprine and xanthinol nicotinate) with placebo (Bhadage 2013 – two arms of a three‐arm trial; Singh 2016). We are uncertain if vasodilators reduce or increase interincisal distance (MD 2.51 mm, 95% CI −0.99 to 6.01; very low‐certainty evidence). There was no heterogeneity (I² = 0%). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial compared intralesional dexamethasone injections to placebo injections (25 participants) Bhadage 2013. The trial included a third arm not used in this analysis. We are uncertain if dexamethasone injections reduce or increase interincisal distance (MD −1.80 mm, 95% CI −6.30 to 2.70; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

Six study subgroups reported outcomes at follow‐up of three to six months (Analysis 2.1).

2.1. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 1: Interincisal distance (mm)

Three trials with 620 participants compared antioxidants with placebo (curcumin, Oxitard and lycopene) (Ara 2018; Arakeri 2020; Patil 2015a). Antioxidants probably increase interincisal distance slightly (MD 8.83 mm, 95% CI 8.22 to 9.45; moderate‐certainty evidence). There was no heterogeneity (I² = 0). We downgraded the certainty of the evidence due to high risk of bias (one level).

Two trials with 136 participants evaluated pentoxifylline. One trial compared pentoxifylline with placebo (Patil 2014a). Prabhu 2015 compared pentoxifylline plus usual care (included intralesional corticosteroid, hyaluronidase and placentrix injections, along with local heat therapy and mouth stretching exercises). We are uncertain whether pentoxifylline increases or decreases interincisal distance (MD −0.15 mm, 95% CI −5.04 to 4.74; very low‐certainty evidence). There was considerable heterogeneity within the subgroup (I² = 92%). We downgraded the certainty of the evidence due to high risk of bias (one level), inconsistency of findings (one level) and imprecision (one level).

We were able to compare two arms of a three‐arm trial, with 25 participants between these two arms (Bhadage 2013). These compared intralesional dexamethasone injections to placebo injections. We are uncertain whether intralesional dexamethasone injections increase or decrease interincisal distance (MD −2.10 mm, 95% CI −6.77 to 2.57; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Two trials with 85 participants compared vasodilators (isoxsuprine and xanthinol nicotinate) with placebo (Bhadage 2013; Singh 2016). We are uncertain whether vasodilators increase or decrease interincisal distance (MD 4.04 mm, 95% CI −1.93 to 10.02; very low‐certainty evidence). There was also substantial heterogeneity (I² = 61%). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

We were able to compare two arms of a three‐arm trial, with 12 participants between these two arms (Cox 2009). These arms compared hydrocortisone plus hyaluronidase with placebo. We are uncertain whether hydrocortisone plus hyaluronidase increases or decreases interincisal distance (MD −4.88 mm, 95% CI −12.87 to 3.11; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

We were able to compare two arms of a three‐arm trial, with 24 participants between these two arms (Cox 2009). These arms compared physiotherapy with no active treatment. We are uncertain whether physiotherapy increases or decreases interincisal distance (MD 0.50 mm, 95% CI −7.72 to 8.72; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

More than six months' follow‐up

One three‐arm trial with 90 participants compared two antioxidants (curcumin, lycopene) versus placebo (Piyush 2018; Analysis 3.1). We combined the antioxidant arms for analysis. Antioxidants may make no difference to interincisal distance at six months' follow‐up or greater (MD −1.41 mm, 95% CI −5.74 to 2.92; low‐certainty evidence). We downgraded the certainty of the evidence due to imprecision (two levels).

3.1. Analysis.

Comparison 3: Any intervention vs placebo (> 6 months' follow‐up), Outcome 1: Interincisal distance (mm)

Burning sensation (VAS 0 to 100 scale)

Four study subgroups reported burning sensation. Of the trials above that reported interincisal distance, only one did not report burning sensation (Cox 2009). Patil 2014a did not report burning sensation using a validated scale and, therefore, was not included in this analysis.

Less than three months' follow‐up

Three study subgroups reported outcomes at follow‐up of less than three months (Analysis 1.2).

1.2. Analysis.

Comparison 1: Any intervention vs placebo (< 3 months' follow‐up), Outcome 2: Burning sensation (VAS)

One trial with 400 participants compared an antioxidant (lycopene) with placebo (Arakeri 2020). Antioxidants probably reduce burning sensation VAS scores (MD −30.92 mm, 95% CI −31.57 to −30.27; moderate‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level).

Two trials with 85 participants compared vasodilators (isoxuprine and xanthinol nicotinate) with placebo (Bhadage 2013 – two arms of a three‐arm trial; Singh 2016). We are uncertain whether vasodilators reduce burning sensation VAS scores (MD −38.79 mm, 95% CI −52.06 to −25.53; very low‐certainty evidence). There was substantial heterogeneity detected within this subgroup (I² = 73%). We downgraded the certainty of the evidence due to high risk of bias (one level), inconsistency (one level) and imprecision (one level).

We were able to compare two arms of a three‐arm trial with 25 participants between these two arms (Bhadage 2013). These arms compared intralesional dexamethasone injections to placebo injections. We are uncertain whether intralesional dexamethasone injections reduce burning sensation VAS scores (MD −52.00 mm, 95% CI −62.04 to −41.96; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

There were four study subgroups that reported outcomes at follow‐up of three to six months (Analysis 2.2), one of which compared the intervention to 'usual care' (Prabhu 2015).

2.2. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 2: Burning sensation (VAS)

Two trials with 500 participants compared antioxidants with placebo (curcumin and lycopene) (Ara 2018; Arakeri 2020). Antioxidants probably reduce burning sensation VAS scores (MD −70.82 mm, 95% CI −94.39 to −47.25; moderate‐certainty evidence). There was considerable heterogeneity (I² = 99%). We downgraded the certainty of the evidence due to high risk of bias (one level); we did not downgrade for heterogeneity as both trials found a clear benefit from the intervention.

One trial with 30 participants compared pentoxifylline tablets with usual care (Prabhu 2015). We are uncertain whether pentoxifylline increases or reduces burning sensation VAS scores (MD −7.00 mm, 95% CI −28.8 to 42.8; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Two trials with 85 participants compared vasodilators (isoxsuprine and xanthinol nicotinate) with placebo (Bhadage 2013 – two arms of a three‐arm trial; Singh 2016). We are uncertain whether vasodilators reduce burning sensation VAS scores (MD −51.02 mm, 95% CI −72.85 to −29.20; very low‐certainty evidence). There was considerable heterogeneity within this subgroup (I² = 91%). We downgraded the certainty of the evidence due to high risk of bias (one level), inconsistency (one level) and imprecision (one level).

We were able to compare two arms of a three‐arm trial, with 25 participants between these two arms (Bhadage 2013). These arms compared intralesional dexamethasone injections to placebo injections. We are uncertain whether intralesional dexamethasone injections reduce burning sensation VAS scores (MD −46.00 mm, 95% CI −53.44 to −38.56; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

More than six months' follow‐up

One three‐arm trial with 90 participants compared antioxidants (curcumin or lycopene) versus placebo (Piyush 2018; Analysis 3.2). We combined the antioxidant arms for analysis. Antioxidants probably reduce burning sensation VAS scores (MD −27.60 mm, 95% CI −36.21 to −18.99; moderate‐certainty evidence). We downgraded the certainty of the evidence due to imprecision (one level).

3.2. Analysis.

Comparison 3: Any intervention vs placebo (> 6 months' follow‐up), Outcome 2: Burning sensation (VAS)

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Head‐to‐head comparisons

There were 22 studies that compared different types of interventions against one another (Alora Veedu 2015; Bhadage 2013; Cox 2009; Dani 2018; Goel 2015; Jiang 2013; Jirge 2008; Kania 2022; Kumar 2007; Mulk 2013; Patil 2014b; Patil 2015b; Patil 2017; Patil 2019; Piyush 2018; Rajbhoj 2021; Singh 2010; Sudarshan 2012; Thakur 2015; Vadepally 2019; Wu 2010; Yadav 2014). The interventions evaluated are summarised in Table 2.

Different types of medications

Participant‐reported resumption of normal eating, chewing and speech

No studies reported this outcome at any time point. One study did report a proxy measure – difficulty in speech and swallowing (Patil 2014b). It reported these as dichotomous outcomes (presence or absence of 'difficulty'). This method of reporting subjective outcomes is not validated and cannot differentiate between degrees of improvement that might represent significant functional gains to study participants; however, for completeness, we presented the results. See Table 5 for a summary of these findings.

4. Additional outcome measure – difficulty in speech, swallowing and mouth opening.

Intervention versus placebo comparisons
Difficulty in speech	2 trials with 226 participants recorded 'difficulty in speech' using a VAS, but then reported this as a dichotomous outcome (presence/absence) (Analysis 2.3). Patil 2015a included 120 participants and compared an antioxidant (Oxitard) with placebo. There was a reduction in the frequency of 'difficulty in speech' (OR 0.51, 95% CI 0.21 to 1.22). In the placebo group, there were 17 participants reporting 'difficulty in speech' compared with 10 participants in the intervention group. We judged this trial at unclear risk of bias. Patil 2014a included 106 participants and compared pentoxifylline with placebo. There was a reduction in the frequency of 'difficulty in speech' (OR 0.20, 95% CI 0.08 to 0.48). In the placebo group, there were 27 participants reporting 'difficulty in speech' compared with 9 participants in the intervention group. We judged this trial at high risk of bias.
Difficulty in swallowing	2 trials with 220 participants compared antioxidants (curcumin and Oxitard) with placebo (Ara 2018; Patil 2015a). There was an overall reduction in the frequency of 'difficulty in swallowing' (OR 0.24, 95% CI 0.10 to 0.60; Analysis 2.4). In the placebo groups, there were 24 participants reporting 'difficulty in swallowing' compared with 7 participants in the intervention groups. Patil 2014a included 106 participants and compared pentoxifylline with placebo. There was a reduction in the frequency of 'difficulty in swallowing' (OR 0.42, 95% CI 0.19 to 0.92). In the placebo group, there were 27 participants reporting 'difficulty in swallowing' compared with 16 participants in the intervention group. We judged this trial at high risk of bias.
Difficulty in mouth opening	1 trial with 100 participants compared an antioxidant to placebo (curcumin) (Ara 2018). There was a reduction in the frequency of participants reporting 'difficulty in mouth opening' (OR 0.00, 95% CI 0.00 to 0.02; Analysis 2.5). In the placebo group, there were 47 participants reporting 'difficulty in mouth opening' compared with 2 in the intervention group. We judged this trial at unclear risk of bias. Ara 2018 reported a reduction in the frequency of participants with 'intolerance to spicy food' (OR 0.02, 95% CI 0.00 to 0.43). In the placebo group, there were 14 participants reporting 'intolerance to spicy food' compared with 0 in the intervention group.
Head‐to‐head comparisons
Difficulty in speech	1 study with 120 participants compared Oxitard capsules with aloe vera gel (Patil 2014b). There was a decrease in the frequency of participants reporting 'difficulty in speech' in favour of Oxitard (OR 0.32, 95% CI 0.14 to 0.76; Analysis 5.3). There were 23 participants reporting 'difficulty in speech' in the aloe vera group compared with 10 participants in the Oxitard group at 3–6 months' follow‐up. This study was at high risk of bias.
Difficulty in swallowing	Patil 2014b reported a decrease in the frequency of participants reporting 'difficulty in swallowing' in favour of Oxitard (OR 0.17, 95% CI 0.07 to 0.44; Analysis 5.4). There were 26 participants reporting 'difficulty in swallowing' in the aloe vera group compared with 7 participants in the Oxitard group at 3–6 months' follow‐up. This study was at high risk of bias.

CI: confidence interval; OR: odds ratio; VAS: visual analogue scale.

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

Six subgroups compared different medical interventions head‐to‐head at follow‐up of less than three months (Analysis 4.1).

4.1. Analysis.

Comparison 4: Medicinal vs alternative medicinal (< 3 months' follow‐up), Outcome 1: Interincisal distance (mm)

Four trials with 128 participants compared intralesional steroid injections with a variety of alternative medicinal treatments. Bhadage 2013 compared intralesional dexamethasone plus hyaluronidase to isoxsuprine tablets; Jiang 2013 compared intralesional triamcinolone to intralesional salvianolic acid B; Kumar 2007 compared intralesional betamethasone plus lycopene tablets to lycopene tablets alone; Yadav 2014 compared intralesional dexamethasone plus hyaluronidase to Turmix tablets. We are uncertain whether intralesional steroid injections reduce or increase interincisal distance (MD −0.64 mm, 95% CI −2.16 to 0.89; low‐certainty evidence). There was no heterogeneity (I² = 0). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

Two trials with 148 participants compared intralesional steroid injections plus adjunctive treatment with intralesion steroid injections alone. Jiang 2013 compared intralesional triamcinolone to intralesional triamcinolone plus salvianolic acid B; Wu 2010 compared intralesional prednisolone acetate to intralesional prednisolone acetate plus salvia miltorrhiza. We are uncertain whether intralesional steroid injections plus adjunctive treatment, compared with intralesion steroid injections alone, reduces or increases interincisal distance (MD −0.03 mm, 95% CI −2.48 to 2.43; low‐certainty evidence). There was no heterogeneity (I² = 0). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

One trial with 30 participants compared levamisole with Antoxid tablets (Jirge 2008). We are uncertain whether levamisole, compared with Antoxid, reduces or increases interincisal distance (MD −1.00 mm, 95% CI −7.18 to 5.18; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 60 participants compared topical curcumin gel with topical aloe vera gel (Rajbhoj 2021). We are uncertain whether topical curcumin gel, compared with aloe vera gel, reduces or increases interincisal distance (MD −0.64 mm, 95% CI −3.91 to 2.63; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 50 participants comparedspirulina with Oxitard (Patil 2019). Spirulina may decrease interincisal distance slightly (MD −2.60 mm, 95% CI −3.82 to −1.38; low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

One trial with 120 participants compared Oxitard with aloe vera gel (Alora Veedu 2015). Oxitard may increase interincisal distance slightly (MD 9.40 mm, 95% CI 8.52 to 10.28; low‐certainty evidence). We downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (one level).

Three to six months' follow‐up

We identified four subgroups comparing different medical interventions head‐to‐head after three to six months' follow‐up (Analysis 5.1).

5.1. Analysis.

Comparison 5: Medicinal vs alternative medicinal (3–6 months' follow‐up), Outcome 1: Interincisal distance (mm)

Three trials with 98 participants compared intralesional steroid injections with a variety of alternative medicinal treatments. The interventions and comparisons are described in the above section (Bhadage 2013; Jiang 2013; Yadav 2014). There may be little to no difference in interincisal distance between treatments (MD −0.57 mm, 95% CI −2.34 to 1.19; low‐certainty evidence). There was no heterogeneity (I² = 0). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

Two trials with 148 participants compared intralesional steroid injections plus adjunctive treatment with intralesion steroid injections alone (Jiang 2013; Wu 2010). There may be little to no difference in interincisal distance between treatments (MD −0.85, 95% CI −3.31 to 1.61; low‐certainty evidence). There was no heterogeneity (I² = 0). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

One trial with 40 participants compared pentoxifylline with spirulina (Mulk 2013). Pentoxifylline may reduce interincisal distance (MD −8.00 mm, 95% CI −14.47 to −1.53; low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (one level).

One trial with 30 participants compared levamisole with Antoxid tablets (Jirge 2008). We are uncertain whether levamisole, compared with Antoxid, reduces or increases interincisal distance (MD −1.00 mm, 95% CI −7.18 to 5.18; very low‐certainty evidence). We downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (two levels).

More than six months' follow‐up

We identified two subgroups comparing medicinal treatments head‐to‐head with more than six months' follow‐up (Analysis 6.1). These subgroups each comprised two arms of the same three‐arm trial (Jiang 2013). We are uncertain whether intralesional triamcinolone, compared with either intralesional salvianolic acid B injections or intralesional triamcinolone plus salvianolic acid B injections, reduces or increases interincisal distance (intralesional salvianolic acid B injections: MD −0.60 mm, 95% CI −3.76 to 2.56; 28 participants; very low‐certainty evidence; intralesional triamcinolone plus salvianolic acid B injections: MD −2.38 mm, 95% CI −5.64 to 0.88; 28 participants; very low‐certainty evidence). For both comparisons, we downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (two levels).

6.1. Analysis.

Comparison 6: Medicinal vs alternative medicinal (> 6 months' follow‐up), Outcome 1: Interincisal distance (mm)

Burning sensation (VAS 0 to 100 scale)

Less than three months' follow‐up

We identified four subgroups comparing different medicinal interventions head‐to‐head with less than three months' follow‐up (Analysis 4.2).

4.2. Analysis.

Comparison 4: Medicinal vs alternative medicinal (< 3 months' follow‐up), Outcome 2: Burning sensation (VAS)

Two trials with 58 participants compared intralesional steroid injections to different alternative medicinal treatments. The interventions and comparisons have been reported in the above section (Bhadage 2013; Jiang 2013). We are uncertain whether steroid injections in comparison to alternative medical treatments reduce or increase burning sensation VAS scores (MD −0.06 mm, 95% CI −10.64 to 10.52; very low‐certainty evidence). There was substantial heterogeneity within this subgroup (I² = 74%). We downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (two levels).

One trial with 28 participants compared intralesional triamcinolone injections to intralesional triamcinolone plus salvianolic acid B injections (Jiang 2013). We are uncertain whether intralesional triamcinolone injections compared to intralesional triamcinolone plus salvianolic acid B injections reduce or increase burning sensation VAS scores (MD −1.00 mm, 95% CI −10.37 to 8.37; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 30 participants compared levamisole versus an antioxidant (Jirge 2008). We are uncertain whether intralesional levamisole, compared with antioxidants, reduces burning sensation VAS scores (MD −9.10 mm, 95% CI −16.23 to −1.97; very low‐certainty evidence). We downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (two levels).

One trial with 60 participants compared topical curcumin gel with aloe vera gel (Rajbhoj 2021). We are uncertain whether topical curcumin gel, compared with aloe vera gel, increases burning sensation VAS scores (MD 16.00 mm, 95% CI 7.21 to 24.79; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

We identified four subgroups comparing different medicinal interventions head‐to‐head with between three and six months' follow‐up (Analysis 5.2).

5.2. Analysis.

Comparison 5: Medicinal vs alternative medicinal (3–6 months' follow‐up), Outcome 2: Burning sensation (VAS)

Three trials with 98 participants compared intralesional steroid injections with a variety of alternative medicinal treatments. The interventions and comparisons are described in the above section (Bhadage 2013; Jiang 2013; Yadav 2014). We are uncertain whether intralesional steroid injections, compared with alternative medicinal treatments, increase or reduce burning sensation VAS scores (MD 8.61 mm, 95% CI −4.52 to 21.74; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level), imprecision (one level) and substantial heterogeneity (I² = 90%).

We were able to compare two arms of a three‐arm trial with 28 participants between these two arms (Jiang 2013). These arms compared intralesional triamcinolone injections with intralesional triamcinolone injections plus salvianolic acid B. We are uncertain whether intralesional triamcinolone injections, compared with intralesional triamcinolone injections plus salvianolic acid B, reduce burning sensation VAS scores (MD −16.30 mm, 95% CI −28.15 to −4.45; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 40 participants compared pentoxifylline to spirulina tablets (Mulk 2013). We are uncertain whether pentoxifylline, compared with spirulina, increases or reduces burning sensation VAS scores (MD 0.50 mm, 95% CI −6.60 to 7.60; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 30 participants compared levamisole versus an antioxidant (Jirge 2008). We are uncertain whether levamisole, compared with antioxidant, reduces burning sensation VAS scores (MD −7.0 mm, 95% CI −13.56 to −0.44; very low‐certainty evidence). We downgraded the certainty of the evidence due to unclear risk of bias (one level) and imprecision (two levels).

More than six months' follow‐up

We identified no subgroups that reported burning sensation VAS scores at more than six months' follow‐up.

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Different surgical techniques

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

We identified two subgroups comparing surgical techniques with one another (Analysis 7.1).

7.1. Analysis.

Comparison 7: Surgery vs alternative surgery (< 3 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with eight participants compared buccal fat pad grafts with extended nasolabial flaps (Patil 2017). We are uncertain if buccal fat pad grafts, compared with extended nasolabial flaps, reduce interincisal distance (MD −7.25 mm, 95% CI −14.19 to −0.31; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with 30 participants compared abdominal dermal fat grafts with nasolabial flaps (Kania 2022). We are uncertain if abdominal dermal fat grafts, compared with nasolabial flaps, reduce interincisal distance (MD −3.60 mm, 95% CI −5.17 to −2.03; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

We identified two subgroups comparing surgical techniques with three to six months' follow‐up data (Analysis 8.1).

8.1. Analysis.

Comparison 8: Surgery vs alternative surgery (3–6 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with 30 participants compared abdominal dermal fat grafts with nasolabial flaps (Kania 2022). We are uncertain if abdominal dermal fat grafts, compared with nasolabial flaps, reduce interincisal distance (MD −14.87 mm, 95% CI −19.93 to −9.81; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

One trial with eight participants compared extended nasolabial flaps to buccal fat pad grafts (Patil 2017). We are uncertain if extended nasolabial flaps, compared with buccal fat pad grafts, reduce or increase interincisal distance (MD −6.00 mm, 95% CI −23.36 to 11.36; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

More than six months' follow‐up

We identified one subgroup that reported interincisal distance at more than six months' follow‐up (Analysis 9.1).

9.1. Analysis.

Comparison 9: Surgery vs alternative surgery (> 6 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with eight participants compared extended nasolabial flaps to buccal fat pad grafts (Patil 2017). We are uncertain if abdominal dermal fat grafts, compared with nasolabial flaps, reduce or increase interincisal distance (MD −6.50 mm, 95% CI −23.72 to 10.72; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Burning sensation (VAS 0 to 100 scale)

No studies comparing surgical modalities measured impact on burning sensation at any time point.

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Surgery alone compared with surgery plus adjunctive treatments

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

We identified one subgroup for this comparison (Analysis 10.1).

10.1. Analysis.

Comparison 10: Surgery + adjunctive medical intervention vs surgery (< 3 months' follow‐up) , Outcome 1: Interincisal distance (mm)

One trial with 10 participants compared buccal fat pad pedicle with placental extract versus buccal fat pad pedicle alone (Thakur 2015). We are uncertain if buccal fat pad pedicle with placental extract, compared with buccal fat pad pedicle alone, increases interincisal distance (MD 7.80 mm, 95% CI 4.90 to 10.70; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

We found no subgroups that evaluated interincisal distance at three to six months' follow‐up for this comparison.

More than six months' follow‐up

We found no subgroups that evaluated interincisal distance at more than six months for this comparison.

Burning sensation (VAS 0 to 100 scale)

No studies comparing surgical modalities with or without adjunctive medical intervention measured burning sensation as an outcome.

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Physiotherapy alone compared with physiotherapy plus adjunctive treatments

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

We identified one subgroup that reported interincisal distance at less than three months' follow‐up (Analysis 11.1).

11.1. Analysis.

Comparison 11: Physiotherapy + adjunctive medical intervention vs physiotherapy (< 3 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with 20 participants compared jaw opening exercises alone to ultrasound plus jaw opening exercises (Dani 2018). We are uncertain if jaw opening exercises alone, compared with ultrasound plus jaw opening exercises, reduce or increase interincisal distance (MD −3.70 mm, 95% CI −11.53 to 4.13; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

We found no subgroups that evaluated interincisal distance at three to six months' follow‐up for this comparison.

More than six months' follow‐up

We found no subgroups that evaluated interincisal distance at more than six months for this comparison.

Burning sensation (VAS 0 to 100 scale)

No studies comparing physiotherapy alone versus physiotherapy plus adjunctive treatments measured burning sensation as an outcome.

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Physiotherapy compared with medication

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

We identified one subgroup that reported interincisal distance at less than three months' follow‐up (Analysis 12.1).

12.1. Analysis.

Comparison 12: Physiotherapy vs medical intervention (3–6 months' follow‐up), Outcome 1: Interincisal distance (mm)

We were able to compare two arms of a three‐arm trial, with 20 participants, comparing tongue depressor stretches to intralesional steroid and hyaluronidase injections (Cox 2009). We are uncertain if tongue depressor stretches, compared with intralesional steroid and hyaluronidase injections, increase or reduce interincisal distance (MD 5.38 mm, 95% CI −2.24 to 13.00; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Three to six months' follow‐up

We found no subgroups that evaluated interincisal distance at three to six months' follow‐up for this comparison.

More than six months' follow‐up

We found no subgroups that evaluated interincisal distance at more than six months for this comparison.

Burning sensation (VAS 0 to 100 scale)

No studies comparing physiotherapy versus medications measured burning sensation as an outcome.

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Surgery combined with different physiotherapy techniques

Maximal mouth opening (interincisal distance (mm))

Less than three months' follow‐up

We found no studies that evaluated interincisal distance at less than three months for this comparison.

Three to six months' follow‐up

We identified one study that reported interincisal distance at three to six months' follow‐up (Analysis 13.1).

13.1. Analysis.

Comparison 13: Physiotherapy vs alternative physiotherapy (3–6 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with 50 participants compared surgery plus Heister jaw opener versus surgery plus wooden tongue depressors (Vadepally 2019). We are uncertain if surgery plus Heister jaw opener, compared with surgery plus wooden tongue depressors, increases interincisal distance (MD 0.92 mm, 95% CI 0.28 to 1.56; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Greater than six months' follow‐up

We identified one study that reported interincisal distance at greater than six months' follow‐up (Analysis 14.1).

14.1. Analysis.

Comparison 14: Physiotherapy vs alternative physiotherapy (> 6 months' follow‐up), Outcome 1: Interincisal distance (mm)

One trial with 50 participants compared surgery plus Heister jaw opener versus surgery plus wooden tongue depressors (Vadepally 2019). We are uncertain if surgery plus Heister jaw opener, compared with surgery plus wooden tongue depressors, increases interincisal distance (MD 0.60 mm, 95% CI 0.01 to 1.19; very low‐certainty evidence). We downgraded the certainty of the evidence due to high risk of bias (one level) and imprecision (two levels).

Burning sensation (VAS 0 to 100 scale)

Vadepally 2019 measured burning sensation on a VAS, but did not fully report the results. The article stated that "regarding burning sensation, there was no significant difference among the 2 groups (p = 0.617)".

Other outcomes

The trials did not measure our other outcomes: range of jaw movement, quality of life, patient satisfaction, hospital admission or costs.

Studies not included in quantitative analysis

We excluded seven studies from meta‐analyses as data were mis‐reported (Mehrotra 2011), or not reported in a usable format (Goel 2015; Kumar 2007; Patil 2015b; Rajendran 2006; Singh 2010; Sudarshan 2012).

Six studies reported important outcomes, such as pain; improvement in intolerance to spicy foods; and reduction in difficulty with eating, swallowing and speaking (Alora Veedu 2015; Ara 2018; Patil 2014a; Patil 2014b; Patil 2015a; Patil 2015b). Table 3 and Table 4 present the findings.

Goel 2015 was a three‐armed trial that compared intralesional betamethasone injections (90 participants) versus lycopene capsules (90 participants) versus placebo (90 participants), and reported data that appeared to show increased interincisal distance in both intervention groups compared to placebo. However, the authors reported only between‐group differences in interincisal distance with no baseline measurements and so changes in interincisal distance could not be determined.

Rajendran 2006 evaluated pentoxifylline tablets (14 participants) compared to 'multivitamins' (15 participants). They reported increased interincisal distance in the pentoxifylline group, but did not report baseline data, so changes in interincisal distance could not be determined. The authors also concluded that the intervention reduced pain. However, pain intensity evaluations were inadequately defined and obtained via "interview data", which makes interpretation of the mean scores presented difficult.

Kumar 2007 was a three‐armed trial that compared oral lycopene (21 participants) versus oral lycopene plus intralesional betamethasone (19 participants) versus placebo (18 participants). They reported data that appeared to show an increase in interincisal distance in the two active intervention groups compared with placebo, but there was limited information provided on withdrawals and incomplete data, which meant the values presented could not be interpreted quantitatively.

Singh 2010 compared intralesional hydrocortisone plus hyaluronidase (50 participants) versus triamcinolone plus hyaluronidase (50 participants). They reported mouth opening as trismus presented categorically rather than as continuous data, which precluded quantitative analysis. The other outcomes reported were not relevant to this review, such as changes in histological appearance of lesions, vesicle formation and burning sensation reported on a non‐validated scale. The authors found no difference between their intervention groups for any outcomes measured.

Sudarshan 2012 investigated topical aloe vera gel (10 participants) versus antioxidant capsules (10 participants). They reported a reduction in burning sensation and improvement in interincisal distance but did not report baseline data, so changes could not be determined.

Adverse effects

Fifteen studies made statements regarding adverse effects. Of these, six did not encounter adverse effects of treatment (Bhadage 2013; Jiang 2013; Patil 2014b; Patil 2019; Singh 2010; Yadav 2014), while nine reported adverse effects in treatment groups (Kania 2022; Kumar 2007; Mulk 2013; Patil 2015b; Patil 2017; Piyush 2018; Singh 2016; Sudarshan 2012; Vadepally 2019).

One study reported three adverse effects, which we classified as serious based on the likelihood of extended hospital stay (Kania 2022, abdominal dermal fat graft versus nasolabial flap). All three adverse effects were graft failure but no further information was provided within the report.

There were no serious adverse events such as death or complications resulting in hospital admission reported following any of the interventions. There were mild and transient general adverse effects of systemic drug use, such as dyspepsia, abdominal pain and bloating, gastritis, and nausea, that may have been linked directly to the interventions (Mehrotra 2011; Mulk 2013; Patil 2015b; Piyush 2018;Sudarshan 2012). One study reported transient peripheral flushing in 90% of participants (Singh 2016). Three participants reported the development of oral ulceration following local and systemic interventions (Kumar 2007; Singh 2016). There were also surgical complications such as infection, wound breakdown (Vadepally 2019), and intraoral hair growth (Patil 2017).

Discussion

Summary of main results

This review assessed the effectiveness of interventions in the management of symptoms and sequelae of OSF. We identified resumption of normal function (e.g. eating, chewing, speech) as being most likely to impact daily activities and, therefore, of the highest priority. We also selected measurement of mouth opening (by interincisal distance) as a proxy measure of function, which has the advantage of being an easily quantifiable and reproducible clinical measure. Severity of burning sensation was also chosen as a frequently reported symptom by people with OSF. Secondary outcomes included quality of life measures and economic considerations associated with treatment. We did not find any studies that reported our secondary outcomes.

This updated review included 30 studies, of which 28 were new. Most studies had methodological issues or were not reported with sufficient information to enable us to completely assess methodological rigour, which limits our confidence in their findings. We judged one trial at overall low risk of bias, five trials at unclear risk and the remaining 24 trials at high risk of bias.

No trials reported our main outcome of interest, which was participant‐reported resumption of normal eating, chewing and speech. Twenty‐nine studies evaluated interincisal distance and 15 evaluated severity of burning sensation using a VAS. Fifteen studies made statements regarding adverse effects. One study reported adverse effects, including serious adverse effects related to surgical procedures. Eight studies reported mild and transient adverse effects of systemic drugs such as dyspepsia, abdominal pain and bloating, gastritis, and nausea. None reported serious adverse events such as death or complications resulting in hospital admission. However, participant numbers were low for each intervention and reporting was inconsistent and lacked detail, so evaluation of intervention safety is severely limited.

We found a wide range of interventions across different trials with several different putative mechanisms of action and modes of delivery. Therefore, we categorised trials into seven broad comparisons: 'any intervention versus placebo', 'different types of medications versus one another', 'different surgical techniques versus one another', 'surgery alone versus surgery plus adjunctive treatments', 'physiotherapy alone versus physiotherapy plus adjunctive treatments', 'physiotherapy versus medications', 'surgery combined with different physiotherapy techniques'. Due to the degree of heterogeneity in interventions, we considered that pooling of subgroups within these comparisons was inappropriate. Therefore, we undertook meta‐analysis at subgroup level only. Often the rationale and clinical justification for the specific interventions and comparisons selected was scant or absent.

None of the studies measured our main outcome 'participant‐reported resumption of normal eating, chewing and speech'. Four studies included a proxy measure that was not validated and provided dichotomous data only (Ara 2018; Patil 2014a; Patil 2014b; Patil 2015a).

We found nine studies with usable data that compared any intervention versus placebo for the outcome of mouth opening measured by interincisal distance in millimetres (Ara 2018; Arakeri 2020; Bhadage 2013; Cox 2009; Patil 2014a; Patil 2015a; Piyush 2018; Prabhu 2015; Singh 2016). Antioxidants may increase mouth opening (indicated by interincisal distance (mm)) when measured at less than three months, and probably increase mouth opening slightly at three to six months. Antioxidants may make no difference to interincisal distance at six months' follow‐up or greater. Pentoxifylline may increase mouth opening slightly at less than three months. It should be noted, however, that these results are all less than 10 mm, which could be considered the minimal change that is meaningful to someone with oral submucous fibrosis. The evidence was very uncertain for all other interventions compared to placebo or no active treatment (intralesional dexamethasone injections, pentoxifylline, hydrocortisone plus hyaluronidase, physiotherapy).

We found six studies with usable data that compared any intervention versus placebo for the outcome of burning sensation (VAS) (Ara 2018; Arakeri 2020; Bhadage 2013; Piyush 2018; Prabhu 2015; Singh 2016). Antioxidants probably reduce burning sensation VAS scores at less than three months, at three to six months and at more than six months (moderate certainty evidence). The evidence was very uncertain for the other interventions that were compared to placebo and measured burning sensation (intralesional dexamethasone, vasodilators).

Fifteen studies reported adverse effects as an outcome. Six of these studies found no adverse effects. One study evaluating abdominal dermal fat graft reported serious adverse effects resulting in prolonged hospital stay for 3/30 participants. There were mild and transient general adverse effects of systemic drugs, such as dyspepsia, abdominal pain and bloating, gastritis, and nausea, in studies evaluating vasodilators and antioxidants in particular. In all the interventions evaluated, the effect of the interventions was small (0.5 mm to 8.83 mm of improvement).

There is no reported MCID in burning sensation VAS scores for people with OSF. One systematic review of MCIDs in chronic pain conditions has suggested 32% relative reduction in VAS scores. As with mouth opening, this needs validating for people with OSF. We found moderate‐certainty evidence that antioxidants reduced burning sensation by 44% at less than three months; 97% at three to six months; 65% at greater than six months.

Overall completeness and applicability of evidence

In this review we presented the evidence regarding a wide variety of interventions in treating people with OSF. There is no recognised standard of care for OSF. The studies included in this review investigated the effectiveness of numerous interventions with different proposed mechanisms of action. Additionally, the 'control' groups often included a clinically active treatment and some studies used combined interventions with overlapping putative mechanisms. This confounds reliable assessment of the effectiveness of either intervention used alone. Therefore, it is difficult to make meaningful comparisons across interventions and studies.

Due to the chronic nature of OSF and the impact of this condition on quality of life, the use of patient‐reported outcome measures (PROMs) reflecting normal daily functional activities, such as eating and speaking, is important. Most trials relied on interincisal distance (29/30 trials), with fewer trials reporting burning sensation (15/30 trials). These outcomes may not reflect functional limitations important to patients, although their assessment is straightforward to administer clinically and relatively reliable. However, only 4/30 trials reported PROMs similar to those prespecified as our primary outcome (participant‐reported resumption of normal eating, chewing and speech). These PROMs were reported in a limited manner (presence or absence of difficulties only).

Adverse effects of treatments should be reported routinely in clinical trials but were omitted in 15/30 trials. Reporting of adverse effects for interventions such as antioxidant supplements is particularly important as these are not subject to the same rigorous safety testing as pharmaceutical products, because they are classified as nutrients by regulatory authorities (Maxwell 1999).

OSF is a chronic condition and any intervention should demonstrate long‐term effectiveness before being considered for routine clinical use. The longest follow‐up period identified with usable data was only nine months. As most studies appear to be evaluating interventions already being used in local clinical practice, longer‐term outcome data beyond the initial trial period should be an important consideration alongside monitoring for adverse effects/events. We found no evidence that this is being undertaken.

Most studies did not report power calculations. Because of the small sample size in many trials, it is likely that studies were not adequately powered to detect differences in treatment if present. Additionally, such small sample sizes can lead to failure to detect rarer but important adverse effects associated with different treatment modalities. Baseline demographics of participants were often not reported, making extrapolation to other patient cohorts challenging. Failure to report baseline demographics meant that we were unable to evaluate the effectiveness of randomisation in balancing groups regarding important confounders, which is especially important in trials with small numbers of participants.

There is no reported MCID in mouth opening for people with OSF. An absolute change of 10 mm has been suggested as the MCID in a temporomandibular joint dysfunction population (Kaur 2022). This needs validating for people with OSF but is the best available estimate at present. Given the overall uncertainty of the evidence, we remain uncertain about the magnitude of the effect estimate. Similarly, there is no reported MCID in burning sensation VAS scores for people with OSF. A systematic review of MCIDs in chronic pain conditions has suggested 32% relative reduction in VAS scores. As with mouth opening, this needs validating for people with OSF. We found moderate‐certainty evidence that antioxidants reduced burning sensation by 44% at less than three months; 97% at three to six months; 65% at greater than six months.

An important part of clinical practice is instruction in cessation of areca nut habit. Most studies included cessation advice as part of their protocol and made attempts to monitor this, but assessment of continued cessation of habit was reported inconsistently. Most protocols included advice to perform mouth stretching exercises alongside the active interventions and so most participants also received some form of physiotherapy, if informally. No trials reported compliance with these ancillary physiotherapy regimens. As most studies were unblinded or inadequately blinded, we also could not rule out the impact of performance bias leading to participants who knew they were receiving an experimental intervention being more compliant with mouth stretching exercises and habit abstinence.

Due to the low certainty of evidence, heterogeneity of interventions and outcomes, and lack of standardised treatment protocols, it is difficult to translate the findings of this review into clinical practice. This is compounded by the use of active interventions as control groups in many studies. Therefore, we find it difficult to justify trials evaluating head‐to‐head comparisons of interventions lacking efficacy data.

Of note, the search identified eight ongoing trials. Seven of these include head‐to‐head comparisons, with only one placebo‐controlled trial. This suggests that the findings of these studies are unlikely to meaningfully change the conclusions of this review.

Quality of the evidence

We included only RCTs in this review, which are considered the gold standard trial design to assess the effectiveness of interventions. However, this depends on the methodological rigour of the studies, which is summarised by risk of bias assessments. Only one of the included trials was at low risk of overall bias (five were unclear and 24 were high). The most common sources of bias were failure to blind participants and selective reporting. Risk of bias judgements were significantly hampered by reporting standards. Almost 25% of judgements were unclear due to insufficient information. Attempts made to clarify information with study authors were largely unsuccessful. Therefore, we consider that trials with unclear risk of bias are likely to reflect a high risk of bias.

Potential biases in the review process

Data from several studies included in this review could not be analysed quantitatively due to inadequate reporting, which prevented data extraction. This could have introduced reporting bias. Where possible, we provided a narrative description of trials with data that did not allow quantitative analysis. We attempted to minimise bias such as obtaining translations where necessary rather than excluding non‐English language papers and contacting authors by email to request missing data. The broad inclusion criteria for types of intervention resulted in a clinically heterogeneous group of trials. Therefore, we considered it was inappropriate to combine subgroups. Consequently, our analyses include small numbers of participants from few trials, which is reflected in our reduced certainty in the evidence. We used Covidence to screen and assess abstracts and full‐text papers for the 2023 update (Covidence). To be included, trials had to employ both a clinical and histological diagnosis for OSF. Although this method ensures that only people with a confirmed diagnosis of OSF are included in this review, it is potentially restrictive as biopsy of suspected OSF lesions is not standard practice in all centres due to the risk of exacerbating symptoms.

Agreements and disagreements with other studies or reviews

There have been several systematic reviews of interventions to manage OSF since the first iteration of this review (Al‐Maweri 2019a; Al‐Maweri 2019b; Awan 2014; Gondivkar 2020; Guo 2020a; Guo 2020b; Guo 2021; Gupta 2020; Kamath 2015; Kerr 2011; More 2020; Rai 2021; Rajesh Kashyap 2021; Xie 2019).

Several reviews of medical interventions broadly mirror our findings. Kerr 2011 reviewed medical interventions and found that the quality and reporting of available research made drawing specific conclusions difficult. More 2020 reviewed innovations in medical management of OSF between 2011 and 2020, encompassing interventions such as steroid injections, hyaluronidase injections, curcumin tablets or gels, Oxitard tablets or pentoxifylline. No individual intervention was effective. The authors concluded that most of the studies reviewed were at high risk of bias and suggested that further high‐quality RCTs with larger sample sizes are required. Awan 2014 also reviewed a wide range of medical interventions to manage OSF; they reported mostly controlled clinical trials and concluded that the evidence was limited to support drug treatments and further high‐quality RCTs are needed. Each review also concluded there is a need for more robust research performed with methodological rigour.

Of the reviews examining medical interventions, three focused specifically on the use of curcumin as an antioxidant in the treatment of OSF (Al‐Maweri 2019a; Guo 2021; Rai 2021). Guo 2021 determined through meta‐analysis that curcumin was not as effective as controls in improving mouth opening and did not reduce oral burning sensation. Al‐Maweri 2019a noted improvement in burning sensation following administration of curcumin, but also reported an inconclusive effect on mouth opening. Rai 2021 reported a small improvement in mouth opening and a reduction in burning sensation following the use of curcumin and turmeric in OSF treatment. Guo 2020b systematically evaluated the antioxidant lycopene in treatment of OSF and found a modest improvement in mouth opening and no difference in alleviation of burning sensation compared with placebo with a short follow‐up period. Gupta 2020 also evaluated lycopene treatments and found no differences between lycopene and other treatments included in their review. One network meta‐analysis examined the effect of herbal derivatives on mouth opening and showed a small improvement with use of lycopene and aloe vera compared with placebo (Rajesh Kashyap 2021). These reviews broadly corroborate our main finding that there may be some benefit to the use of antioxidants such as lycopene in the treatment of OSF. They mirrored our concerns about the reliability of the evidence due to the low number of RCTs and high risk of bias. These reviews did not identify any serious adverse effects of antioxidant interventions, but found limited reporting of adverse effects/events in the included studies. There are several putative adverse effects associated with specific antioxidants, including overdose, toxicity and drug interactions alongside those noted in our review. More rare instances of adverse effects have been reported, but the available evidence is mostly limited to case reports and series (Maxwell 1999). To our knowledge, there have been no systematic safety evaluations of the antioxidant supplements used in the included studies.

Liu 2018 undertook a systematic review and meta‐analysis to evaluate the efficacy of pentoxifylline for treatment of OSF. They included three studies and found a modest improvement in interincisal distance. All studies had a low number of participants and there was considerable heterogeneity between them. The same authors also found that the effectiveness of the intervention was greatest at less than one month but did not persist longer term. Similarly, there was reduction in burning sensation that was sustained beyond three months of treatment. The adverse effects they reported corroborated our findings and included gastrointestinal disturbances and anxiety. These adverse effects were reported to be dose‐dependent. No serious adverse effects/events were reported. The authors concluded that pentoxifylline is an effective short‐ and long‐term treatment for OSF, although they noted a lack of accepted treatments for the condition. We consider that the conclusions about the long‐term effectiveness of pentoxifylline were not supported by the findings of their review, and we found that pentoxifylline does not appear to improve mouth opening at three to six months.

One systematic review of aloe vera in the management of OSF combined six trials in meta‐analyses and reported an improvement in burning sensation with topical application of aloe vera at one and two months of follow‐up (Al‐Maweri 2019b). There was no difference between intervention and control after three months of follow‐up. There were no differences in mouth opening noted. The included trials had high heterogeneity; five were judged at high risk of bias, and one unclear. Similarly, we found no evidence to support the effectiveness of aloe vera for burning sensation or mouth opening in people with OSF.

We did not find sufficient evidence to support or refute the effectiveness of hyaluronidase combined with corticosteroids in the management of OSF. This finding corroborates a systematic review and meta‐analysis by Guo 2020a, which combined six studies with 244 participants and also concluded that hyaluronidase with corticosteroids had no benefit over control drugs in improving mouth opening or alleviating burning sensation. However, it should be noted that the included studies had small numbers of participants and mostly made head‐to‐head comparisons with other drugs purported to manage the symptoms of OSF. There were no placebo‐controlled trials included. The one placebo‐controlled study we found showed no effectiveness of this intervention.

One review of surgical interventions for OSF found a lack of controlled studies, small numbers of participants and short follow‐up periods (Kamath 2015). There was often no standardisation of treatment protocols. Most studies were case reports or series and the authors concluded that the effectiveness of treatments could not be assessed, noting that surgeon preference was often the factor that determined treatment approach. This review echoes our findings; the small number of randomised surgical trials we identified had low numbers of participants and were at high risk of bias.

Gondivkar 2020 reviewed mouth exercising devices and identified five studies that examined five different devices with the goal of improving or maintaining mouth opening. Although one of these trials was an RCT and was included in our review (Cox 2009), we judged it at high risk of bias due to blinding and allocation issues. The remaining devices were not evaluated by controlled trials and so, although the authors reported improvements in mouth opening for most of the devices, these findings should be interpreted cautiously. It is also of note that the studies had short follow‐up periods. It is unlikely that mouth stretching devices have any serious deleterious effects, but we agree with the authors' conclusions that more methodologically rigorous research is required.

Authors' conclusions

Implications for practice.

Based on the quality of the available evidence and wide range of interventions evaluated, with multiple potential biological mechanisms of action, it is difficult to draw any firm conclusions about treatment of oral submucous fibrosis (OSF). Any treatment provided should be preceded by carefully evaluating the potential risks and benefits, with highly invasive treatments such as surgical intervention, which carry a high risk of morbidity, only used with clear justification on an individual case basis.

Within the above limitations, we found moderate‐certainty evidence that antioxidants administered systemically probably improve mouth opening slightly at three to six months and improve burning sensation VAS scores up to and beyond six months. The clinical importance of the improvements in mouth opening are unclear; the improvements in burning sensation may be clinically important based on findings from chronic pain literature. We found only low/very low‐certainty evidence for all other comparisons and outcomes. There was insufficient evidence to make an informed judgement about potential adverse effects associated with any of these treatments. There was insufficient evidence to support or refute the effectiveness of the other interventions tested.

Based on the findings, antioxidant supplements show promise in the management of OSF. However, prescribers should be cautious and consider potential drug interactions and ensure patients understand dose recommendations, as many supplements are available without prescription. It is also important for clinicians to report any adverse events – expected or unexpected – that may be associated with antioxidant use, such as by using the UK Medicines and Healthcare products Regulatory Agency's (MHRA) 'Yellow Card' scheme. Except for the effect of antioxidants on burning sensation, there was insufficient evidence to determine the effectiveness of interventions for longer than six months. Given the lack of evidence around effective treatments, the mainstay approach to OSF will continue to be prevention using behavioural interventions to support habit cessation.

Implications for research.

High‐quality intervention trials with a low risk of bias that compare treatments for OSF against non‐active controls are needed. Future trials should be conducted and reported according to the CONSORT statement (www.consort-statement.org/), and focus on biologically plausible interventions with a proposed mechanism of action identified and tested from studies conducted in vitro or using animal models. In addition, future trials should standardise research methods, in particular focusing on patient‐reported outcomes and standardising assessment of other outcomes such as interincisal distance and burning sensation. It would be useful to establish agreed minimum clinically important differences for key outcomes. Investment in understanding the pathogenesis of OSF and interventions to prevent the disease by educating and bringing about a shift in cultural and behavioural attitudes around areca nut consumption is needed.

What's new

Date	Event	Description
28 February 2024	New search has been performed	Searches updated to September 2022
28 February 2024	New citation required and conclusions have changed	New authors. 28 new included studies

History

Protocol first published: Issue 2, 2008
Review first published: Issue 4, 2008

Appendices

Appendix 1. Cochrane Oral Health's Trials Register search strategy

Cochrane Oral Health's Trials Register is available via the Cochrane Register of Studies. For information on how the register was compiled, see oralhealth.cochrane.org/trials.

1 MESH DESCRIPTOR Oral Submucous Fibrosis AND INREGISTER
2 ((submucous AND fibrosis) AND (oral or mouth*)) AND INREGISTER
3 "oral submucous fibrosis" AND INREGISTER
4 #1 or #2 or #3

Appendix 2. Cochrane Central Register of Controlled Clinical Trials (CENTRAL) search strategy

#1 ORAL SUBMUCOUS FIBROSIS/
#2 ((submucous NEXT fibrosis) AND (oral or mouth))
#3 #1 or #2

Appendix 3. MEDLINE Ovid search strategy

1 ORAL SUBMUCOUS FIBROSIS/

2 (“submucous fibrosis” AND (oral or mouth))

3 OR/1‐2

The above subject search was linked with the highly sensitive search strategy designed by Cochrane for identifying randomised controlled trials and controlled clinical trials in MEDLINE (as described in Lefebvre 2022).

1. randomized controlled trial.pt.
2. controlled clinical trial.pt.
3. randomized.ab.
4. placebo.ab.
5. drug therapy.fs.
6. randomly.ab.
7. trial.ab.
8. groups.ab.
9. or/1‐8
10. exp animals/ not humans.sh.
11. 9 not 10

Appendix 4. Embase Ovid search strategy

(“submucous fibrosis” AND (oral or mouth))

The above subject search was linked with the highly sensitive search strategy designed by Cochrane for identifying randomised controlled trials and controlled clinical trials in Embase (as described in Lefebvre 2022).

1. Randomized controlled trial/

2. Controlled clinical study/

3. random$.ti,ab.

4. randomization/

5. intermethod comparison/

6. placebo.ti,ab.

7. (compare or compared or comparison).ti.

8. ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab.

9. (open adj label).ti,ab.

10. ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab.

11. double blind procedure/

12. parallel group$1.ti,ab.

13. (crossover or cross over).ti,ab.

14. ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab.

15. (assigned or allocated).ti,ab.

16. (controlled adj7 (study or design or trial)).ti,ab.

17. (volunteer or volunteers).ti,ab.

18. human experiment/

19. trial.ti.

20. or/1‐19

21. random$ adj sampl$ adj7 ("cross section$" or questionnaire$1 or survey$ or database$1)).ti,ab. not (comparative study/ or controlled study/ or randomi?ed controlled.ti,ab. or randomly assigned.ti,ab.)

22. Cross‐sectional study/ not (randomized controlled trial/ or controlled clinical study/ or controlled study/ or randomi?ed controlled.ti,ab. or control group$1.ti,ab.)

23. (((case adj control$) and random$) not randomi?ed controlled).ti,ab.

24. (Systematic review not (trial or study)).ti.

25. (nonrandom$ not random$).ti,ab.

26. "Random field$".ti,ab.

27. (random cluster adj3 sampl$).ti,ab.

28. (review.ab. and review.pt.) not trial.ti.

29. "we searched".ab. and (review.ti. or review.pt.)

30. "update review".ab.

31. (databases adj4 searched).ab.

32. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$1).ti. and animal experiment/

33. Animal experiment/ not (human experiment/ or human/)

34. or/21‐33

35. 20 not 34

Appendix 5. US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov) search strategy

submucous fibrosis

Appendix 6. World Health Organization International Clinical Trials Registry Platform search strategy

submucous fibrosis

Appendix 7. Outcomes not relevant for this review

Studies reported several other outcomes that we did not deem reliable or relevant to this review:

• improvement in mucosa tightness measured by visual analogue scale (Alora Veedu 2015);

• presence of a uvula deformity (Ara 2018);

• presence of mucosal blanching (Ara 2018; Patil 2014b);

• change in tongue protrusion (Ara 2018; Kania 2022; Kumar 2007; Mulk 2013; Patil 2019; Piyush 2018; Prabhu 2015; Rajendran 2006; Singh 2010; Singh 2016; Sudarshan 2012; Yadav 2014);

• changes in various histological scores (Bhadage 2013; Singh 2010);

• change in serum levels of immunoglobulin (Ig)G, IgA and IgM (Jirge 2008);

• change in lesion size (Patil 2015a; Patil 2019; Wu 2010);

• various measures of cheek flexibility (Patil 2019; Piyush 2018; Rajendran 2006; Singh 2010; Singh 2016; Sudarshan 2012);

• duration of surgery (Kania 2022);

• scarring (Kania 2022);

• thickness of masseter (Kania 2022).

Data and analyses

Comparison 1. Any intervention vs placebo (< 3 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Interincisal distance (mm)	5		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.1.1 Antioxidants vs placebo	2	520	Mean Difference (IV, Random, 95% CI)	3.11 [0.46, 5.77]
1.1.2 Pentoxifylline vs placebo	1	106	Mean Difference (IV, Random, 95% CI)	1.80 [1.02, 2.58]
1.1.3 Vasodilators vs placebo	2	85	Mean Difference (IV, Random, 95% CI)	2.51 [‐0.99, 6.01]
1.1.4 Intralesional dexamethasone injections vs placebo	1	25	Mean Difference (IV, Random, 95% CI)	‐1.80 [‐6.30, 2.70]
1.2 Burning sensation (VAS)	3		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.2.1 Antioxidant vs placebo	1	400	Mean Difference (IV, Random, 95% CI)	‐30.92 [‐31.57, ‐30.27]
1.2.2 Vasodilator vs placebo	2	85	Mean Difference (IV, Random, 95% CI)	‐38.79 [‐52.06, ‐25.53]
1.2.3 Dexamethasone vs placebo	1	25	Mean Difference (IV, Random, 95% CI)	‐52.00 [‐62.04, ‐41.96]

Comparison 2. Any intervention vs placebo (3–6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Interincisal distance (mm)	8		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.1.1 Antioxidant vs placebo	3	620	Mean Difference (IV, Random, 95% CI)	8.83 [8.22, 9.45]
2.1.2 Pentoxifylline vs placebo	2	136	Mean Difference (IV, Random, 95% CI)	‐0.15 [‐5.04, 4.74]
2.1.3 Dexamethasone vs placebo	1	25	Mean Difference (IV, Random, 95% CI)	‐2.10 [‐6.77, 2.57]
2.1.4 Vasodilator vs placebo	2	85	Mean Difference (IV, Random, 95% CI)	4.04 [‐1.93, 10.02]
2.1.5 Hydrocortisone + hyaluronidase vs placebo	1	12	Mean Difference (IV, Random, 95% CI)	‐4.88 [‐12.87, 3.11]
2.1.6 Physiotherapy vs control	1	24	Mean Difference (IV, Random, 95% CI)	0.50 [‐7.72, 8.72]
2.2 Burning sensation (VAS)	5		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.2.1 Antioxidant vs placebo	2	500	Mean Difference (IV, Random, 95% CI)	‐70.82 [‐94.39, ‐47.25]
2.2.2 Pentoxifylline vs usual care	1	30	Mean Difference (IV, Random, 95% CI)	7.00 [‐28.79, 42.79]
2.2.3 Vasodilator vs placebo	2	85	Mean Difference (IV, Random, 95% CI)	‐51.02 [‐72.85, ‐29.20]
2.2.4 Dexamethasone vs placebo	1	25	Mean Difference (IV, Random, 95% CI)	‐46.00 [‐53.44, ‐38.56]
2.3 Difficulty in speech	2		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
2.3.1 Pentoxifylline vs placebo	1	106	Odds Ratio (M‐H, Random, 95% CI)	0.20 [0.08, 0.48]
2.3.2 Antioxidant vs placebo	1	120	Odds Ratio (M‐H, Random, 95% CI)	0.51 [0.21, 1.22]
2.4 Difficulty in swallowing	3		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
2.4.1 Pentoxifylline vs placebo	1	106	Odds Ratio (M‐H, Random, 95% CI)	0.42 [0.19, 0.92]
2.4.2 Antioxidant vs placebo	2	220	Odds Ratio (M‐H, Random, 95% CI)	0.24 [0.10, 0.60]
2.5 Difficulty in mouth opening	1		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
2.5.1 Antioxidant vs placebo	1	100	Odds Ratio (M‐H, Random, 95% CI)	0.00 [0.00, 0.02]
2.6 Intolerance to spicy food	1		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
2.6.1 Antioxidant vs placebo	1	100	Odds Ratio (M‐H, Random, 95% CI)	0.02 [0.00, 0.43]

2.3. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 3: Difficulty in speech

2.4. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 4: Difficulty in swallowing

2.5. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 5: Difficulty in mouth opening

2.6. Analysis.

Comparison 2: Any intervention vs placebo (3–6 months' follow‐up), Outcome 6: Intolerance to spicy food

Comparison 3. Any intervention vs placebo (> 6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
3.1.1 Curcumin or lycopene vs placebo	1	90	Mean Difference (IV, Random, 95% CI)	‐1.41 [‐5.74, 2.92]
3.2 Burning sensation (VAS)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
3.2.1 Antioxidant vs placebo	1	90	Mean Difference (IV, Random, 95% CI)	‐27.60 [‐36.21, ‐18.99]

Comparison 4. Medicinal vs alternative medicinal (< 3 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Interincisal distance (mm)	9		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.1.1 Steroid vs alternative (medicinal treatment)	4	128	Mean Difference (IV, Random, 95% CI)	‐0.64 [‐2.16, 0.89]
4.1.2 Steroid vs alternative (steroid + adjunct)	2	148	Mean Difference (IV, Random, 95% CI)	‐0.03 [‐2.48, 2.43]
4.1.3 Levamisole vs alternative (antioxidant)	1	30	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐7.18, 5.18]
4.1.4 Topical curcumin gel vs alternative (topical aloe vera gel)	1	60	Mean Difference (IV, Random, 95% CI)	‐0.64 [‐3.91, 2.63]
4.1.5 Spirulina vs alternative (oxitard)	1	50	Mean Difference (IV, Random, 95% CI)	‐2.60 [‐3.82, ‐1.38]
4.1.6 Oxitard capsules vs alternative (aloe vera gel)	1	120	Mean Difference (IV, Random, 95% CI)	9.40 [8.52, 10.28]
4.2 Burning sensation (VAS)	4		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.2.1 Steroid vs alternative	2	58	Mean Difference (IV, Random, 95% CI)	‐0.06 [‐10.64, 10.52]
4.2.2 Steroid vs alternative (steroid + adjunct)	1	28	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐10.37, 8.37]
4.2.3 Levamisole vs alternative (antioxidant)	1	30	Mean Difference (IV, Random, 95% CI)	‐9.10 [‐16.23, ‐1.97]
4.2.4 Topical curcumin gel vs alternative (topical aloe vera gel)	1	60	Mean Difference (IV, Random, 95% CI)	16.00 [7.21, 24.79]

Comparison 5. Medicinal vs alternative medicinal (3–6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Interincisal distance (mm)	6		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.1.1 Steroid vs alternative	3	98	Mean Difference (IV, Random, 95% CI)	‐0.57 [‐2.34, 1.19]
5.1.2 Steroid vs alternative (steroid + adjunct)	2	148	Mean Difference (IV, Random, 95% CI)	‐0.85 [‐3.31, 1.61]
5.1.3 Pentoxifylline vs alternative (spirulina)	1	40	Mean Difference (IV, Random, 95% CI)	‐8.00 [‐14.47, ‐1.53]
5.1.4 Levamisole vs alternative (antioxidant)	1	30	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐7.18, 5.18]
5.2 Burning sensation (VAS)	5		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.2.1 Steroid vs alternative	3	98	Mean Difference (IV, Random, 95% CI)	8.61 [‐4.52, 21.74]
5.2.2 Steroid vs alternative (steroid + adjunct)	1	28	Mean Difference (IV, Random, 95% CI)	‐16.30 [‐28.15, ‐4.45]
5.2.3 Pentoxifylline vs alternative (spirulina)	1	40	Mean Difference (IV, Random, 95% CI)	0.50 [‐6.60, 7.60]
5.2.4 Levamisole vs alternative (antioxidant)	1	30	Mean Difference (IV, Random, 95% CI)	‐7.00 [‐13.56, ‐0.44]
5.3 Difficulty in speech	1		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
5.3.1 Antioxidant vs topical aloe vera	1	120	Odds Ratio (M‐H, Random, 95% CI)	0.32 [0.14, 0.76]
5.4 Difficulty in swallowing	1		Odds Ratio (M‐H, Random, 95% CI)	Subtotals only
5.4.1 Antioxidant vs topical aloe vera	1	120	Odds Ratio (M‐H, Random, 95% CI)	0.17 [0.07, 0.44]

5.3. Analysis.

Comparison 5: Medicinal vs alternative medicinal (3–6 months' follow‐up), Outcome 3: Difficulty in speech

5.4. Analysis.

Comparison 5: Medicinal vs alternative medicinal (3–6 months' follow‐up), Outcome 4: Difficulty in swallowing

Comparison 6. Medicinal vs alternative medicinal (> 6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
6.1.1 Steroids vs alternative	1	28	Mean Difference (IV, Random, 95% CI)	‐0.60 [‐3.76, 2.56]
6.1.2 Steroid vs alternative (steroid + adjunct)	1	28	Mean Difference (IV, Random, 95% CI)	‐2.38 [‐5.64, 0.88]

Comparison 7. Surgery vs alternative surgery (< 3 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Interincisal distance (mm)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
7.1.1 Buccal fat pad vs alternative (extended nasolabial flap)	1	8	Mean Difference (IV, Random, 95% CI)	‐7.25 [‐14.19, ‐0.31]
7.1.2 Abdominal dermal fat graft vs alternative (nasolabial flap)	1	30	Mean Difference (IV, Random, 95% CI)	‐3.60 [‐5.17, ‐2.03]

Comparison 8. Surgery vs alternative surgery (3–6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Interincisal distance (mm)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.1.1 Abdominal dermal fat graft vs alternative (nasolabial flap)	1	30	Mean Difference (IV, Random, 95% CI)	‐14.87 [‐19.93, ‐9.81]
8.1.2 Extended nasolabial flap vs alternative (buccal fat pad)	1	8	Mean Difference (IV, Random, 95% CI)	‐6.00 [‐23.36, 11.36]

Comparison 9. Surgery vs alternative surgery (> 6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only

Comparison 10. Surgery + adjunctive medical intervention vs surgery (< 3 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only

Comparison 11. Physiotherapy + adjunctive medical intervention vs physiotherapy (< 3 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
11.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
11.1.1 Ultrasound + jaw opening exercises vs jaw opening exercises	1	20	Mean Difference (IV, Random, 95% CI)	‐3.70 [‐11.53, 4.13]

Comparison 12. Physiotherapy vs medical intervention (3–6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
12.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
12.1.1 Tongue depressor stretches vs Local injection hyaluronidase + steroid	1	20	Mean Difference (IV, Random, 95% CI)	5.38 [‐2.24, 13.00]

Comparison 13. Physiotherapy vs alternative physiotherapy (3–6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
13.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
13.1.1 Heister jaw opener vs wooden tongue depressors	1	50	Mean Difference (IV, Random, 95% CI)	0.92 [0.28, 1.56]

Comparison 14. Physiotherapy vs alternative physiotherapy (> 6 months' follow‐up).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
14.1 Interincisal distance (mm)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
14.1.1 Heister jaw opener vs wooden tongue depressors	1	50	Mean Difference (IV, Random, 95% CI)	0.60 [0.01, 1.19]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Alora Veedu 2015.

Study characteristics
Methods	Study design: 3‐arm parallel‐group RCT
Participants	Baseline characteristics Group I (hyaluronidase submucosal injection) Initial participant number randomised to group: 15 Group II (dexamethasone submucosal injection) Initial participant number randomised to group: 15 Group III (combination of hyaluronidase and dexamethasone submucosal injection) Initial participant number randomised to group: 15 Overall Initial participant number: 45 Inclusion criteria All clinically diagnosed and histopathologically confirmed OSF cases reported at the outpatient department of Oral Medicine and Radiology from March 2011 to February 2012 Exclusion criteria Severe systemic medical problems History of drug allergy or hypersensitivity to hyaluronidase, dexamethasone or lignocaine Treated for OSF or for malignancy arising in OSF Unwilling to participate Pregnancy Temporomandibular problems or pericoronitis of the mandibular third molars
Interventions	Intervention characteristics Hyaluronidase submucosal injection (Group I) Description of intervention: submucosal injections of hyaluronidase 1500 IU, bilateral buccal mucosa, biweekly for 5 weeks Follow‐up period: 6 months Concurrent treatment: massage cheeks, mouth opening physiotherapy 5 times daily and passive tongue spatula placement between teeth Dexamethasone submucosal injection (Group II) Description of intervention: submucosal injections of dexamethasone 2 mL (4 mg/mL) with 1 mL of 2% lignocaine with adrenaline (1:80,000), bilateral buccal mucosa, biweekly for 5 weeks Follow‐up period: 6 months Concurrent treatment: massage cheeks, mouth opening physiotherapy 5 times daily and passive tongue spatula placement between teeth Combination of hyaluronidase and dexamethasone submucosal injection (Group III) Description of intervention: submucosal injections of a combination of dexamethasone 1 mL (4 mg/mL) and hyaluronidase (750 IU) diluted with 1 mL of 2% lignocaine with adrenaline (1:80,000) and 1 mL of water for injection. Bilateral buccal mucosa, biweekly for 5 weeks Follow‐up period: 6 months Concurrent treatment: massage cheeks, mouth opening physiotherapy 5 times daily and passive tongue spatula placement between teeth
Outcomes	Burning sensation improvement: VAS Pain while opening mouth improvement: VAS Tightness of mucosa improvement: VAS Interincisal mouth opening: mm
Identification	Sponsorship source: State Board of Medical Research, Kerala, India Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomly divided into three groups of 15 each by the permuted block randomization method."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Similar interventions, with allocation concealment. Allocation performed by co‐investigator.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Co‐investigator outcome collection, blinded until after statistical analysis.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Co‐investigator outcome collection, blinded until after statistical analysis.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Attrition and exclusions not reported, final population not reported in results.
Selective reporting (reporting bias)	Low risk	All outcomes fully reported.
Other bias	Low risk	No other sources of bias identified.

Ara 2018.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group I (placebo) Initial participant number randomised to group: 50 Group II (curcumin) Initial participant number randomised to group: 50 Overall Initial participant number: 100 Inclusion criteria People with clinical stage 2 OSF with clinically and histopathologically confirmed diagnosis Exclusion criteria Clinical stage 1 and 3 OSF Oromucosal disorders with clinical features same as OSF Under treatment Clinically diagnosed cases not ready for incisional biopsy Medically compromised conditions
Interventions	Placebo (Group I) Description of intervention: 2 placebo capsules daily (duration unknown) Follow‐up period: unknown Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis. Intervention started 1 month after cessation of habit Curcumin (Group II) Description of intervention: 2 × curcumin 500 mg capsules daily (duration unknown) Follow‐up period: unknown Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis. Intervention started 1 month after cessation of habit
Outcomes	Difficulty in mouth opening: number reporting presence of difficulty Burning sensation: number reporting presence of burning sensation Intolerance to spicy food: number reporting presence of intolerance Difficulty in swallowing: number reporting presence of difficulty Uvula: number with each change present (shrunken; hockey stick deformity) Hockey stick uvula: number with this present Blanching of soft tissues: number with this present at each site (buccal mucosa; labial mucosa; hard palate; soft palate; floor of mouth; tongue) Fibrosis of soft tissues: number with this present at each site (buccal mucosa; labial mucosa; retromolar area; soft palate) Burning sensation: VAS Pain: VAS Interincisal mouth opening: mm Tongue protrusion: mm Cheek flexibility: change in distance between 2 reference points on cheek inflation Clinical stage I–III: number with each stage Histopathological grade I–IV: number each grade
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used simple random sampling technique.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Used placebo capsules.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to impact results.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Low risk	All outcomes reported.
Other bias	Low risk	No concerns.

Arakeri 2020.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group A (lycopene) Initial participant number randomised to group: 200 Group B (placebo) Initial participant number randomised to group: 200 Overall Initial participant number: 400 Inclusion criteria Clinical symptoms of Group II OSF (early cases with interincisal openings of 26–35 mm) Exclusion criteria Known chronic illnesses Allergy to any of the products used in the study Previous history of operations for OSF Secondary oral changes present
Interventions	Intervention characteristics Lycopene (Group A) Description of intervention: lycopene 4 mg tablets twice daily for 3 months Follow‐up period: 6 months Concurrent treatment: habit cessation advice and mouth opening physiotherapy Placebo (Group B) Description of intervention: placebo tablets orally twice daily for 3 months Follow‐up period: 6 months Concurrent treatment: habit cessation advice and mouth opening physiotherapy
Outcomes	Burning sensation improvement: VAS Interincisal mouth opening: mm
Identification	Sponsorship source: none reported Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make a judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make a judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Used placebo.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low rate of loss to follow‐up.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Bhadage 2013.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Isoxsuprine (Group A) Initial participant number randomised to group: 15 Dexamethasone (Group B) Initial participant number randomised to group: 15 Placebo (Group C) Initial participant number randomised to group: 10 Overall Initial participant number: 45 Inclusion criteria History of oral burning sensation when eating Subjectively observed gradual reduction in mouth opening Feeling of loss of elasticity in buccal tissues Bilateral buccal fibrous bands observed objectively on palpation by the investigator Histological diagnosis of OSF by biopsy of buccal mucosa Exclusion criteria Oral malignancy Other mucosal diseases (including chronic ulcers, leukoplakia, lichen planus and infectious mucositis) Trismus due to other conditions Earlier treatment for oral submucous fibrosis Systemic diseases Regular medication Debilitated by age Pregnant women
Interventions	Intervention characteristics Group A (isoxsuprine) Description of intervention: isoxsuprine 10 mg tablet orally, 4 times daily for 6 weeks Follow‐up period: 4 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy and cheek blowing exercises. Habit cessation advice Group B (dexamethasone) Description of intervention: bilateral intralesional injection of dexamethasone 2 mL with hyaluronidase 1500 IU mixed with 1 mL of 2% lidocaine with adrenaline 1:80,000, twice weekly Duration of intervention: 6 weeks Follow‐up period: 4 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy and cheek blowing exercises. Habit cessation advice Group C (placebo) Description of intervention: tablets containing fine sugar orally for 6 weeks Follow‐up period: 4 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy and cheek blowing exercises. Habit cessation advice
Outcomes	Interincisal mouth opening: mm Burning sensation: VAS Histological scoring of pre‐ and postintervention biopsy: severity score 0–4
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Clear difference between interventions (tablet vs injection) and placebo differs from both interventions (capsule).
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not effectively blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	Notably higher attrition in groups A and C. No intention‐to‐treat analysis.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Cox 2009.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Control Initial participant number randomised to group: 16 Hyaluronidase and steroid injections Initial participant number randomised to group: 15 Physiotherapy Initial participant number randomised to group: 23 Overall Initial participant number: 54 Inclusion criteria OSF confirmed by biopsy Clinical evidence of mucosal fibrosis Subjective reduction in oral opening Objective reduction in oral opening (below 32 mm incisal distance) Exclusion criteria Oral squamous cell carcinoma People with surgical treatment planned due to severe restriction of oral opening
Interventions	Intervention characteristics Control Description of intervention: untreated for 4 months Follow‐up period: none Concurrent treatment provided alongside intervention: habit cessation advice, diet advice and correction of any identified underlying anaemia (mebendazole 100 mg twice daily for 3 days, and ferrous sulphate 200 mg and folic acid 5 mg both daily for 4 months) Hyaluronidase and steroid injections Description of intervention: injection of hyaluronidase 1500 units and hydrocortisone 100 mg in sterile water, given biweekly for 4 weeks Follow‐up period: none Concurrent treatment provided alongside intervention: habit cessation advice, diet advice and correction of any identified underlying anaemia (mebendazole 100 mg twice daily for 3 days, and ferrous sulphate 200 mg and folic acid 5 mg both daily for 4 months) Physiotherapy Description of intervention: tongue spatulas passively placed between anterior teeth, holding jaws open for 1 minute, performed 5 times per session with 5 sessions per day for 4 months. An additional spatula was added every 5th day unless pain (in which case delayed to 10th day) Follow‐up period: none Concurrent treatment provided alongside intervention: habit cessation advice, diet advice and correction of any identified underlying anaemia (mebendazole 100 mg twice daily 3 days, and ferrous sulphate 200 mg and folic acid 5 mg both daily for 4 months)
Outcomes	Interincisal mouth opening improvement: mm
Identification	Sponsorship source: not declared Country: Nepal
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Random number allocation not applied universally – participants unable to attend for injections were assigned interventions alternately.
Allocation concealment (selection bias)	High risk	Allocation was not concealed – participants reassigned to physiotherapy group if unable to attend for injections.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind participants. Personnel not reported.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Incomplete outcome data (attrition bias) All outcomes	High risk	High attrition.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Dani 2018.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Experimental group (ultrasound with jaw exercises) Initial participant number randomised to group: 13 Control (jaw exercises only) Initial participant number randomised to group: 10 Overall Initial participant number: 23 Inclusion criteria Both males and females Ages 16 to 40 years Confirmed diagnosis of OSF referred for physiotherapy Confirmation of histopathological diagnosis received from author by email Exclusion criteria Reduced mouth opening due to other problems than OSF Patient declining to participate Patient participating in any other ongoing research project
Interventions	Intervention characteristics Ultrasound with jaw exercises (experimental group) Description of intervention: ultrasound therapy for 7 minutes for 1 week (intervals unclear) Follow‐up period: unclear Concurrent treatment provided alongside intervention: jaw opening exercises (maintain maximal mouth opening for 5 seconds, lateral deviations and protrusion of mandible, repeated 10 times) for 1 week Jaw exercises only (control) Description of intervention: none Follow‐up period: unclear Concurrent treatment provided alongside intervention: jaw opening exercises (maintain maximal mouth opening for 5 seconds, lateral deviations and protrusion of mandible, repeated 10 times) for 1 week
Outcomes	Interincisal mouth opening: unclear
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomization was done using shuffling cards and sequentially numbered."
Allocation concealment (selection bias)	Low risk	Quote: "sealed, opaque envelopes were used as a method of concealment."
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind interventions due to obvious differences in intervention nature.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Outcome assessor was blinded.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Outcome assessor was blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low attrition rate.
Selective reporting (reporting bias)	Low risk	All outcomes fully reported.
Other bias	Low risk	No concerns.

Goel 2015.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Control Initial participant number randomised to group: 90 Betamethasone injection Initial participant number randomised to group: 90 Lycopene capsule Initial participant number randomised to group: 90 Overall Initial participant number: 270 Inclusion criteria Aged 20–50 years Complaining of difficulty in mouth opening History of consumption of areca nut in various forms Palpable fibrotic bands with blanched oral mucosa Complaining of ulceration or burning sensation Biopsy‐confirmed diagnosis of OSF Exclusion criteria Unwilling to participate Diagnosed with systemic conditions such as anaemia Edentulous OSF superimposed by any other lesion Complaining of difficulty in opening the mouth due to other reasons (inflammation, infection, trauma, temporomandibular joint disorders) > 1 chewing habit
Interventions	Intervention characteristics Control Description of intervention: no treatment Follow‐up period: 12 months Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis Betamethasone injection Description of intervention: intralesional injection of betamethasone 4 mg/mL diluted in 1.0 mL of 2% xylocaine given in buccal mucosa (centrally in regions of third molar, first molar and first premolar) bilaterally (half dose on each side), administered biweekly for 6 months Follow‐up period: 12 months Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis Lycopene capsule Description of intervention: lycopene 2 mg capsules, twice daily for 6 months Follow‐up period: 12 months Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis
Outcomes	Interincisal mouth opening improvement: mm
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind participants. Personnel not reported.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to make judgement.
Selective reporting (reporting bias)	High risk	Outcomes collected not specified in methods.
Other bias	Low risk	No concerns.

Jiang 2013.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group A (triamcinolone) Initial participant number randomised to group: unknown Group B (salvianolic acid B) Initial participant number randomised to group: unknown Group C (combination of triamcinolone and salvianolic acid B) Initial participant number randomised to group: unknown Overall Initial participant number: 42 Inclusion criteria Clinical diagnosis of OSF (betel nut chewing history; burning sensation in the mouth; xerostomia; blanching; stiffness of oral mucosa; fibrous bands in buccal mucosa; progressive inability to open the mouth) Biopsy‐confirmed diagnosis of OSF Male or female Age 20–50 years Able to follow the investigator's arrangement and sign informed consent form voluntarily Agreement not to use other drugs to cure OSF during the procedure OSF located in the cheek or pterygomandibular regions only Exclusion criteria Allergy to Salvia miltiorrhizae Other mucosal disorders (e.g. leukoplakia, lichen planus or erythroplakia) System disorders (i.e. diabetes mellitus, myasthenia gravis, AIDS) Any treatment for OSF before entering the current study Taking systemic non‐steroidal anti‐inflammatory drugs or immunomodulatory agents within the past 2 weeks Pregnancy or lactation
Interventions	Intervention characteristics Triamcinolone (Group A) Description of intervention: intralesional injection of triamcinolone 2 mg after 5 minutes of local anaesthetic cream application at weekly intervals for 20 weeks Follow‐up period: 24 weeks Concurrent treatment provided alongside intervention: mouth stretching exercises 3 times a day for 15 minutes and habit cessation advice Salvianolic acid B (Group B) Description of intervention: intralesional injection of salvianolic acid B 4 mg after 5 minutes of local anaesthetic cream application at weekly intervals for 20 weeks Follow‐up period: 24 weeks Concurrent treatment provided alongside intervention: mouth stretching exercises 3 times a day for 15 minutes and habit cessation advice Combination of triamcinolone and salvianolic acid B (Group C) Description of intervention: intralesional injection of triamcinolone 2 mg and salvianolic acid B 4 mg after 5 minutes of local anaesthetic cream application, at weekly intervals for 20 weeks Follow‐up period: 24 weeks Concurrent treatment provided alongside intervention: mouth stretching exercises 3 times a day for 15 minutes and habit cessation advice
Outcomes	Interincisal mouth opening improvement: mm Burning sensation: change in value of VAS
Identification	Sponsorship source: 'Scientific Research Project of Xiangnan University', 'Scientific Research Project of the First People's Hospital of Chenzhou', 'University of South China', and 'Science Technology Bureau Research Project of Chanzhou Municipal' Country: China
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Group C received 2 injections, Groups A and B received 1 injection.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Quote: "Maximum mouth opening (interincisal distance) was measured by one operator who was blind to the groups."
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low attrition.
Selective reporting (reporting bias)	Low risk	All outcomes reported.
Other bias	Low risk	No concerns.

Jirge 2008.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group I (levamisole) Initial participant number randomised to group: 15 Group II (antoxid) Initial participant number randomised to group: 15 Group III (combination of levamisole and antoxid) Initial participant number randomised to group: 15 Overall Initial participant number: 45 Inclusion criteria Aged ≥ 15 years Apparently healthy and well oriented in time, space and as a person Satisfies the characteristic clinical features of OSF with histopathological confirmation and had burning sensation Not taking any medication for their disease condition Agreed to the biopsy, blood and immunological examination Willing for follow‐up visits Willing to quit the habit of chewing areca nut, gutkha and tobacco in any form Exclusion criteria People with OSF and any past or present systemic diseases (e.g. diabetes, hypertension, liver disorders or kidney diseases, autoimmune disorders) Other mucosal lesions With acute or chronic infection Known allergy or contraindication to the study drugs
Interventions	Intervention characteristics Levamisole (Group A) Description of intervention: levamisole 50 mg tablet orally 3 times daily for 3 consecutive days in a week, for 3 alternate weeks Follow‐up period: 60 days Concurrent treatment provided alongside intervention: cessation of pan/gutkha habit Antoxid (Group B) Description of intervention: 1 capsule antoxid twice daily for 6 weeks Follow‐up period: 60 days Concurrent treatment provided alongside intervention: cessation of pan/gutkha habit Combination of levamisole and antoxid (Group C) Description of intervention: levamisole 50 mg tablet orally 3 times daily for 3 consecutive days in a week, for 3 alternate weeks. Along with 1 capsule antoxid twice daily for 6 weeks Follow‐up period: 60 days Concurrent treatment provided alongside intervention: cessation of pan/gutkha habit
Outcomes	Interincisal opening: mm Burning sensation: VAS Serum IgG, IgA and IgM: levels before and after treatment
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Insufficient information to make judgement.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to make judgement.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes fully reported.
Other bias	Unclear risk	No baseline group characteristics reported.

Kania 2022.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group I (abdominal dermal fat graft) Initial participant number randomised to group: 15 Group II (nasolabial flap) Initial participant number randomised to group: 15 Overall Initial participant number: 30 Inclusion criteria Grade 3 and 4A OSF (Khanna and Andrade classification) Mouth opening < 25 mm Biopsy‐confirmed OSF High levels of motivation to give up deleterious habits and perform aggressive post‐operative mouth opening exercises Exclusion criteria Grade 1 and 2 OSF Biopsy report suggestive of malignancy
Interventions	Intervention characteristics Abdominal dermal fat graft (Group I) Description of intervention: excision of fibrous bands, mouth opening using Heister's mouth gag (35–40 mm), coronoidectomy, reconstruction with abdominal dermal fat pad Follow‐up period: 6 months Concurrent treatment provided alongside intervention: extraction of third molars. Intensive physiotherapy using Heister's mouth gag 5–6 times daily for 20–30 minutes each time 3 days postoperatively onwards Nasolabial flap (Group II) Description of intervention: excision of fibrous bands, mouth opening using Heister's mouth gag (35–40 mm), coronoidectomy, reconstruction with nasolabial flap Follow‐up period: 6 months Concurrent treatment provided alongside intervention: extraction of third molars. Intensive physiotherapy using Heister's mouth gag 5–6 time daily for 20–30 minutes each time 3 days postoperatively onwards
Outcomes	Interincisal mouth opening: mm Time taken for mucosalisation: weeks Intercommissure distance: mm Duration of surgery: hours Graft uptake: frequency Complications: frequency Scarring: Stony Brook Scar Evaluation Scale Patient satisfaction score: Likert scale Thickness of masseter: cm
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computer assisted random sequence generator."
Allocation concealment (selection bias)	High risk	Unethical to conceal allocation prior to surgery.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not possible to blind participants or personnel due to obvious differences in procedures.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Blinding not possible.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported fully.
Other bias	Low risk	No other concerns.

Kumar 2007.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Group A (oral lycopene) Initial participant number randomised to group: 21 Group B (oral lycopene and intralesional betamethasone) Initial participant number randomised to group: 19 Group C (placebo capsules) Initial participant number randomised to group: 18 Overall Initial participant number: 58 Inclusion criteria History of chewing areca nut Difficulty in swallowing and chewing Burning pain on eating spicy food Restricted mouth opening changes in the oral mucous membrane (palpable vertical fibrous bands, stiffness, blanching) Histopathological confirmation Exclusion criteria None specified
Interventions	Oral lycopene (Group A) Description of intervention: lycopene 16 mg orally daily in 2 divided doses for 2 months Follow‐up period: 6 months Concurrent treatment: habit cessation advice Oral lycopene and intralesional betamethasone (Group B) Description of intervention: lycopene 16 mg orally daily in 2 divided doses and intralesional injections of betamethasone (2 × 1 mL ampules of 4 mg each) twice‐weekly for 2 months Follow‐up period: 6 months Concurrent treatment: habit cessation advice Placebo capsules (Group C) Description of intervention: placebo capsules for 2 months Follow‐up period: 6 months Concurrent treatment: habit cessation advice
Outcomes	Improvement in interincisal opening: mm Improvement in tongue protrusion: units of 5 mm Presence of palpable fibrous bands in the buccal mucosa: present or absent Burning sensation: present, persisting, reduced or absent
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind as groups underwent different interventions (injection vs capsule).
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	High attrition rate.
Selective reporting (reporting bias)	High risk	All predetermined outcomes reported but not at all time points.
Other bias	Low risk	No concerns.

Mehrotra 2011.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Group A (placebo) Initial participant number randomised to group: 30 Group B (pentoxifylline) Initial participant number randomised to group: 32 Overall Initial participant number: 62 Inclusion criteria People reporting difficulty in chewing and restricted mouth opening Presence of fibrous bands Histopathologically confirmed diagnosis of OSF Exclusion criteria People who refused scalpel biopsy People with medical problems People with dental appliances (e.g. orthodontic or fixed prostheses) that could potentially interfere with examination
Interventions	Intervention characteristics Placebo (Group A) Description of intervention: multivitamin therapy (similar in packaging and taste to pentoxifylline), twice daily for 30 days, then 3 times daily for 6 months Follow‐up period: 18 months Concurrent treatment provided alongside intervention: habit cessation advice Pentoxifylline (Group B) Description of intervention: pentoxifylline 400 mg, twice daily for 30 days, then 3 times daily for 6 months Follow‐up period: 18 months Concurrent treatment provided alongside intervention: habit cessation advice
Outcomes	Interincisal mouth opening: unknown Burning sensation: unknown Adverse events: percentage occurrence
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and investigators blinded.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Investigators blinded.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to make judgement.
Selective reporting (reporting bias)	High risk	Multiple outcomes and time points not reported.
Other bias	Low risk	No concerns.

Mulk 2013.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Pentoxifylline (Group 1) Initial participant number randomised to group: 20 Spirulina (Group 2) Initial participant number randomised to group: 20 Overall Initial participant number: 40 Inclusion criteria Clinically and histopathologically diagnosed cases of OSF Exclusion criteria People with hypertension, diabetes mellitus, anaemic stomatitis, radiation fibrosis, scleroderma, immunosuppressive diseases, peptic ulcer bleeding disorders, cardiac disorders People who had received treatment for oral submucous fibrosis before People with hypersensitivity to study drugs
Interventions	Intervention characteristics Pentoxifylline (Group 1) Description of intervention: oral pentoxifylline 400 mg twice daily for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: mouth opening exercises daily Spirulina (Group 2) Description of intervention: oral spirulina capsules 500 mg twice daily Follow‐up period: 3 months Concurrent treatment provided alongside intervention: mouth opening exercises daily
Outcomes	Interincisal mouth opening: cm Burning sensation: VAS Tongue protrusion: cm
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Simple randomisation method."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants not blinded.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	10 participants excluded (20% of initial sample) at end of study for poor compliance. Unclear which group. No intention‐to‐treat analysis.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes fully reported.
Other bias	Low risk	No concerns.

Patil 2014a.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Pentoxifylline (Group A) Initial participant number randomised to group: 53 Placebo (Group B) Initial participant number randomised to group: 53 Overall Initial participant number: 106 Inclusion criteria People with clinically and histopathologically diagnosed OSF reporting to the Department of Oral Medicine and Radiology Male or female Exclusion criteria People with any evidence of severe psychiatric, cardiac, gastrointestinal or metabolic disorders, and pregnancy and lactation
Interventions	Intervention characteristics Pentoxifylline (Group A) Description of intervention: pentoxifylline 400 mg orally twice daily for 3 months Follow‐up period: 2 months Concurrent treatment provided alongside intervention: habit cessation advice Placebo (Group B) Description of intervention: multivitamin orally for 3 months (unknown interval) Follow‐up period: 2 months Concurrent treatment provided alongside intervention: habit cessation advice
Outcomes	Interincisal mouth opening: mm Tongue protrusion: mm Difficulty in speech – number reporting absence, reduction or presence of difficulty: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Difficulty in swallowing – number reporting absence, reduction or presence of difficulty: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) Burning sensation – number reporting absence, reduction or presence of burning sensation: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) Pain associated with lesion – number reporting absence, reduction or presence of pain: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present)
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Placebo used but unclear if investigators were aware of group allocation.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No attrition.
Selective reporting (reporting bias)	High risk	Not all outcomes fully reported.
Other bias	Low risk	No concerns.

Patil 2014b.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Oxitard (Group A) Initial participant number randomised to group: 60 Aloe vera (Group B) Initial participant number randomised to group: 60 Overall Initial participant number: 120 Inclusion criteria Diagnosed clinically and histopathologically for OSF Male or female Not taking any medications for OSF previously Exclusion criteria Any evidence of hypersensitivity to aloe vera Severe psychiatric, cardiac, gastrointestinal or metabolic disorders, pregnancy and lactation
Interventions	Intervention characteristics Oxitard (Group A) Description of intervention: 2 Oxitard capsules orally twice daily for 3 months Follow‐up period: 2 months Concurrent treatment provided alongside intervention: habit cessation advice Aloe Vera (Group B) Description of intervention: aloe vera gel 5 mg applied topically 3 times daily for 3 months Follow‐up period: 2 months Concurrent treatment provided alongside intervention: habit cessation advice
Outcomes	Interincisal mouth opening: mm Tongue protrusion: mm Pain associated with lesion – number reporting absence, reduction or presence of pain: absent, present or reduced Difficulty in swallowing – number reporting absence, reduction or presence of difficulty: absent, present or reduced Difficulty in speech – number reporting absence, reduction or presence of difficulty: absent, present or reduced Change in size of lesion: area of blanched mucosa
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded – different intervention modalities.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition reported as (quote) "no dropouts from the study due to any reason."
Selective reporting (reporting bias)	High risk	Burning sensation only partially reported. Size of lesion only partially reported.
Other bias	Low risk	No concerns.

Patil 2015a.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Oxitard (Group A) Initial participant number randomised to group: 60 Placebo (Group B) Initial participant number randomised to group: 60 Overall Initial participant number: 120 Inclusion criteria Clinico‐pathologically diagnosed OSF Male or female Exclusion criteria With any evidence of severe psychiatric, cardiac, gastrointestinal or metabolic disorders, and pregnancy and lactation
Interventions	Intervention characteristics Oxitard (Group A) Description of intervention: 2 Oxitard capsules twice daily for 3 months Follow‐up period: stopped when intervention ceased Concurrent treatment provided alongside intervention: none Placebo (Group B) Description of intervention: placebo tablets orally twice daily for 3 months Follow‐up period: stopped when intervention ceased Concurrent treatment provided alongside intervention: none
Outcomes	Interincisal mouth opening: mm Tongue protrusion: mm Difficulty in speech – number reporting absence, reduction or presence of difficulty: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Difficulty in swallowing – number reporting absence, reduction or presence of difficulty: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Burning sensation – number reporting absence, reduction or presence of burning sensation: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Pain associated with lesion – number reporting absence, reduction or presence of pain: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Change in size of lesion: mild (< 1.5 cm), moderate (1.5–2.5 cm) or severe (> 2.5 cm) decrease in size
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Placebo controlled.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "None of the patients withdrew from the study due to any reason."
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Patil 2015b.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Spirulina (Group A) Initial participant number randomised to group: 21 Aloe vera (Group B) Initial participant number randomised to group: 21 Overall Initial participant number: 42 Inclusion criteria People with clinico‐pathologically diagnosed OSF reporting to the Department of Oral Medicine and Radiology Males amid females Exclusion criteria With any evidence of severe psychiatric, cardiac, gastrointestinal or metabolic disorders, and pregnancy and lactation
Interventions	Intervention characteristics Spirulina (Group A) Description of intervention: spirulina 250 mg orally twice daily for 3 months Follow‐up period: 2 months Concurrent treatment provided alongside intervention: none Aloe vera (Group B) Description of intervention: aloe vera gel 5 mg topically 3 times daily for 3 months Follow‐up period: 2 months Concurrent treatment provided alongside intervention: none
Outcomes	Interincisal mouth opening: mm Burning sensation – number reporting absence, reduction or presence of burning sensation: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Pain associated with lesion – number reporting absence, reduction or presence of pain: absent, present or reduced (VAS used: 0–1 = absent; 1–6 = reduced; 7–10 = present) (score ranges as reported in publication) Ulcers/vesicles/erosion – number with absence, reduction or presence of lesions
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Interventions different – capsule vs gel.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "There were no dropouts from the study due to any reason during the follow‐up."
Selective reporting (reporting bias)	High risk	Cheek flexibility not reported, difficulty with swallowing/speech not reported.
Other bias	Low risk	No concerns.

Patil 2017.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Extended nasolabial flaps (Group 1) Initial participant number randomised to group: 4 Buccal fat pad graft (Group 2) Initial participant number randomised to group: 4 Overall Initial participant number: 8 Inclusion criteria People who reported to department with clinical diagnosis of oral submucous fibrosis Moderately advanced or advanced stage of OSF with interincisal mouth opening ≤ 15 mm and with longstanding positive history of habits Diagnosis is confirmed clinically, histopathologically and any malignancy excluded High level of motivation to give up deleterious habits Exclusion criteria Medically compromised (ASA class III and IV) Unwilling to give consent and or attend 1‐year follow‐up History of surgical treatment for oral submucous fibrosis Clinical or radiological presentation of any other cause of trismus
Interventions	Intervention characteristics Extended nasolabial flaps (Group 1) Description of intervention: procedure under general anaesthetic and bilaterally. Incision and release of fibrous bands. Coronoidectomy. Optimum mouth opening held with Fergusson's mouth gag. Reconstruction with extended nasolabial flap. Procedure performed once only Follow‐up period: 12 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy for 6 months postoperatively Buccal fat pad graft (Group 2) Description of intervention: procedure under general anaesthetic and bilaterally. Incision and release of fibrous bands. Coronoidectomy. Optimum mouth opening held with Fergusson's mouth gag. Reconstruction with buccal fat pad graft. Procedure performed once only Follow‐up period: 12 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy for 6 months postoperatively
Outcomes	Interincisal mouth opening: mm
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind the participants. Personnel not reported.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No attrition.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Patil 2019.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Spirulina (Group I) Initial participant number randomised to group: 25 Oxitard (Group II) Initial participant number randomised to group: 25 Overall Initial participant number: 50 Inclusion criteria Clinico‐pathologically diagnosed OSF cases seen within the Department of Oral Medicine and Radiology, Saveetha Dental College and Hospitals, Chennai, India Exclusion criteria Evidence of severe psychiatric, cardiac, gastrointestinal or metabolic disorders Pregnancy, lactation or both
Interventions	Intervention characteristics Spirulina (Group I) Description of intervention: spirulina 500 mg in 2 divided doses for 3 months Follow‐up period: 3 months Concurrent treatment: habit cessation Oxitard (Group II) Description of intervention: 2 Oxitard capsules twice daily for 3 months Follow‐up period: 3 months Concurrent treatment: habit cessation
Outcomes	Interincisal mouth opening: mm Tongue protrusion: mm Cheek flexibility: mm Presence/absence/reduction of ulcers, erosions, vesicles and burning sensation: VAS
Identification	Sponsorship source: none reported Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	States "simple randomisation" but no further details provided.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding reported.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	No blinding, participant‐reported outcome.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Low risk	No concerns.
Other bias	Low risk	No other concerns.

Piyush 2018.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Curcumin (Group A) Initial participant number randomised to group: 30 Lycopene (Group B) Initial participant number randomised to group: 30 Placebo (Group C) Initial participant number randomised to group: 30 Overall Initial participant number: 90 Inclusion criteria People with clinical characteristics of OSF (areca nut habit, blanching, palpable fibrous bands) with histopathological confirmation No evidence of dysplasia or malignancy histopathologically No history of prior treatment for the same Has quit the chewing habit and accepted regular follow‐up Has mouth opening of 15–35 mm Exclusion criteria Did not give consent for participation Cardiac, respiratory, renal and hepatic diseases Pregnant women and lactating mothers Presence of other premalignant lesion or white patch (e.g. leukoplakia, oral lichen planus) History of hypersensitivity to turmeric and other antioxidants
Interventions	Intervention characteristics Curcumin (Group A) Description of intervention: 1 Turmix tablet (combination of curcumin and piperine 300 mg) twice daily for 6 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Lycopene (Group B) Description of intervention: 1 capsule of lycopene 8 mg twice daily for 6 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Placebo (Group C) Description of intervention: 1 capsule of placebo once daily for 6 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none
Outcomes	Interincisal mouth opening: mm Burning sensation: VAS Tongue protrusion: mm Cheek flexibility: change in distance between 2 reference points on cheek inflation Interincisal mouth opening improvement: change in mm Burning sensation improvement: change in value of VAS Tongue protrusion improvement: change in mm Cheek flexibility improvement: change in cheek flexibility measurements
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Simple random sampling was performed using computer‐generated randomization to distribute patients into three groups."
Allocation concealment (selection bias)	Low risk	Quote: "Method of concealment was sequentially numbered, sealed, opaque envelopes."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Both the participant and outcome assessor were blinded."
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Assessors blinded. Use of placebo.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "All participants completed treatment and reported."
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported.
Other bias	Low risk	No concerns.

Prabhu 2015.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Pentoxifylline and conventional therapies (Group I) Initial participant number randomised to group: 15 Conventional therapies only (Group II) Initial participant number randomised to group: 15 Overall Initial participant number: 30 Inclusion criteria Clinically confirmed cases of oral submucous fibrosis Habitual chewers of areca‐quid or pan masala Aged > 18 years Exclusion criteria History of any systemic disease such as diabetes, hypertension Pregnant or lactating mothers Have previously exhibited intolerance to pentoxifylline or other xanthines such as caffeine or theophylline Aged < 18 years Any developmental defects or TMJ disorders
Interventions	Intervention characteristics Pentoxifylline and conventional therapies (Group I) Description of intervention: pentoxifylline 2 × tablets daily for 15 days and then increased to 3 × tablets daily if no adverse effects Follow‐up period: 4 months Concurrent treatment provided alongside intervention: conventional therapies included intralesional corticosteroid, hyaluronidase and placentrix injections, along with local heat therapy and mouth stretching exercises Conventional therapies only (Group II) Description of intervention: clinician‐directed usual care Follow‐up period: 3 months Concurrent treatment provided alongside intervention: conventional therapies included intralesional corticosteroid, hyaluronidase and placentrix injections, along with local heat therapy and mouth stretching exercises
Outcomes	Interincisal mouth opening: mm Burning sensation: VAS Tongue protrusion: mm Microvascular density: unclear Area percentage occupied by blood vessels: percentage
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No evidence of blinding.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No attrition.
Selective reporting (reporting bias)	High risk	Not all histological and patient‐reported outcomes reported.
Other bias	Low risk	No concerns.

Rajbhoj 2021.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Curcumin gel (Group A) Initial participant number randomised to group: 30 Aloe vera (Group B) Initial participant number randomised to group: 30 Overall Initial participant number: 60 Inclusion criteria Diagnosed clinically and histopathologically for OSF Having had no OSF treatment for last 3–4 months Ready for regular follow‐up Willing to quit habits Exclusion criteria Hypersensitive to aloe vera Hypersensitive to curcumin Suffering from psychiatric disorder Having any TMJ dysfunction OSF with malignant transformation
Interventions	Intervention characteristics Curcumin gel (Group A) Description of intervention: daily topical application of curcumin gel 5 mg applied in 3–4 divided doses Follow‐up period: 6 weeks Concurrent treatment: mouth opening physiotherapy 5 times a day Aloe vera gel (Group B) Description of intervention: daily topical application of aloe vera gel5 mg applied in 3–4 divided doses Follow‐up period: 6 weeks Concurrent treatment: mouth opening physiotherapy 5 times a day
Outcomes	Burning sensation improvement: VAS Interincisal mouth opening: mm
Identification	Sponsorship source: none reported Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Simple randomisation technique."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and personnel not blinded.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported.
Other bias	Low risk	No concerns.

Rajendran 2006.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Pentoxifylline (group I) Initial participant number randomised to group: 14 Multivitamins (group II) Initial participant number randomised to group: 15 Overall Initial participant number: 29 Inclusion criteria Clinically advanced stage oral submucous fibrosis Habitual chewers of areca‐quid Exclusion criteria Systemic disease including hypertension, diabetes, cardiac disease, duodenal and gastric ulcers, bleeding dyscrasias
Interventions	Intervention characteristics Pentoxifylline (Group I) Description of intervention: pentoxifylline 2 × 400 mg tablets orally daily for 30 days and then increased to 3 × tablets daily if no adverse effects Follow‐up period: 7 months Concurrent treatment provided alongside intervention: 1 multivitamin tablet at night, local heat therapy and mouth opening exercises Multivitamins (group II) Description of intervention: none Follow‐up period: 7 months Concurrent treatment provided alongside intervention: 1 multivitamin tablet at night, local heat therapy and mouth opening exercises
Outcomes	Interincisal mouth opening improvement: change in mm Tongue protrusion improvement: change in mm Relief from fibrotic bands: improvement or no improvement Relief from intolerance to spicy food: improvement or no improvement Relief from burning sensation: improvement or no improvement Improvement in salivation, rigidity of the mucosa and depapillation of dorsum tongue: improvement or no improvement Improvement from tinnitus: improvement or no improvement Improvement from difficulty in swallowing, speech: improvement or no improvement
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded – experimental treatment group underwent extensive investigation.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Attrition not reported.
Selective reporting (reporting bias)	High risk	Not all outcomes reported at all time points.
Other bias	Low risk	No concerns.

Singh 2010.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Hydrocortisone and hyaluronidase (Group A) Initial participant number randomised to group: 50 Triamcinolone and hyaluronidase (Group B) Initial participant number randomised to group: 50 Overall Initial participant number: 100 Inclusion criteria Patients of stage II OSF having trismus Trismus defined as mouth opening less than normal (< 5.25 cm in males and < 4.75 cm in females) Exclusion criteria Stage I and III
Interventions	Intervention characteristics Hydrocortisone and hyaluronidase (Group A) Description of intervention: injection of a combination of hydrocortisone acetate (1.5 mL of 25 mg/mL) and hyaluronidase 1500 IU submucosally in retromolar trigone, soft palate and cheek bilaterally (half dose on each side). Performed every 15 days for 22 weeks Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Triamcinolone and hyaluronidase (Group B) Description of intervention: injection of a combination of triamcinolone acetonide (10 mg/ml) and hyaluronidase 1500 IU submucosally in retromolar trigone, soft palate and cheek bilaterally (half dose on each side). Performed every 15 days for 22 weeks Frequency of intervention: every 15 days Duration of intervention: 22 weeks Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none
Outcomes	Burning sensation with spicy food: subjective score 0–10 Burning sensation with spicy food improvement: reduction in subjective score Ulcers/vesicles repeated formation: subjective score 0–10 Ulcers/vesicles repeated formation: reduction in subjective score Trismus: objective score of 0 (males > 5.25 cm; females > 4.75 cm), 2 (males 3–5.25 cm; females 3–4.75 cm), 5 (2–3 cm) or 10 (< 2 cm) Trismus improvement: reduction in objective score Ankyloglossia: objective score of 5 if partial protrusion and 10 if no protrusion Ankyloglossia improvement: reduction in objective score Ulcers/vesicles present: objective score of 1 (unilateral, single lesion), 2 (bilateral, single lesions), 3 (unilateral, multiple lesions) or 4 (bilateral, multiple lesions) Ulcers/vesicles improvement: reduction in objective score Fibrosis: objective scores of 2 for presence in each area (soft palate and uvula; anterior faucial pillars and tonsils; buccal mucosae and gingivobuccal sulci; retromolar trigones; tongue; floor of mouth) Fibrosis: reduction in objective score Total (sign + symptom) score: sum of overall scores for signs and symptoms Total (sign + symptom) score: reduction in sum of overall scores Histopathological score: histopathological stages of 'very early', 'early', 'moderately advanced', and 'advanced', all assigned scores of 1, 2, 3 and 4, respectively Histopathological score improvement: reduction in histopathological score Histopathological stage numbers: number at each stage of 'early', 'moderately advanced', and 'advanced'
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised by lottery system into either group A or B. Chosen from 50 chits/tickets for each group.
Allocation concealment (selection bias)	High risk	No allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants blinded. Unclear personnel blinding but standardised technique used.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Quote: "Assessed by another observer. The other clinician who was observing response to treatment was not aware of the treatment group."
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "All the patients who were registered were followed up for 3 months after completion of treatment. There were no dropouts."
Selective reporting (reporting bias)	Low risk	All prespecified outcomes are reported.
Other bias	Low risk	No concerns.

Singh 2016.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Xanthinol nicotinate injection (Group I) Initial participant number randomised to group: 30 Saline injection (Group II) Initial participant number randomised to group: 30 Overall Initial participant number: 60 Inclusion criteria People clinically and histopathologically diagnosed with OSF between December 2012 and March 2014 Exclusion criteria Pregnancy History of recent myocardial infarction, cerebrovascular accidents, severely compromised cardiac function Caffeine or theophylline intolerance Hypersensitive to the drug 'xanthinol nicotinate' Positive drug history of anticoagulants
Interventions	Intervention characteristics Xanthinol nicotinate injection (Group I) Description of intervention: intralesional injections xanthinol nicotinate biweekly for 4 months. Follow‐up period: 4 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy Saline injection (Group II) Description of intervention: intralesional saline injections biweekly for 4 months Follow‐up period: 4 months Concurrent treatment provided alongside intervention: mouth opening physiotherapy
Outcomes	Interincisal mouth opening: mm, reported individually for OSF clinical stages I–IV Burning sensation: VAS, reported individually for OSF clinical stages I–IV Cheek flexibility: change in distance between 2 reference points on cheek inflation, reported individually for OSF clinical stages I–IV Tongue protrusion: mm, reported individually for OSF clinical stages I–IV
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Placebo injections used.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	Low risk	Participants blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No participants lost to follow‐up.
Selective reporting (reporting bias)	Low risk	All outcomes reported.
Other bias	Low risk	No concerns.

Sudarshan 2012.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Aloe vera (Group A) Initial participant number randomised to group: 10 Antioxidant (Group B) Initial participant number randomised to group: 10 Overall Initial participant number: 20 Inclusion criteria OSF diagnosed clinically and histopathologically No previous treatment for OSF Ready to quit the habit of gutka chewing and accepted for regular follow‐ups OSF mouth opening 20–39 mm Exclusion criteria Coexisting systemic illness Other mucosal lesions (e.g. leukoplakia, malignant lesions) History of hypersensitivity to aloe vera Pregnant women and lactating mothers Severe cases (< 20 mm mouth opening)
Interventions	Intervention characteristics Aloe vera (Group A) Description of intervention: aloe vera gel 5 mg applied topically to buccal mucosa bilaterally 3 times daily for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: habit cessation advice Antioxidant (Group B) Description of intervention: antioxidant capsules (beta‐carotene 5 mg; mixed carotenoids 5 mg; vitamin E 25 mg; chromium 200 μg; zinc 7.5 mg; manganese 1.5 mg; copper 1 mg; selenium 150 μg; vitamin C 100 mg) orally twice daily for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: habit cessation advice
Outcomes	Burning sensation: VAS Burning sensation improvement: change in value of VAS Interincisal mouth opening: mm Interincisal mouth opening improvement: change from baseline in mm Tongue protrusion: mm Tongue protrusion improvement: change from baseline in mm Cheek flexibility: change in distance between 2 reference points on cheek inflation Cheek flexibility improvement: change from baseline measurement
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible to blind due to differences in interventions.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "There were no drop outs during the follow‐up period."
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported.
Other bias	Low risk	No concerns.

Thakur 2015.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Surgery and placental extract (Group S) Initial participant number randomised to group: 5 Surgery (Group C) Initial participant number randomised to group: 5 Overall Initial participant number: 10 Inclusion criteria ASA grade I Interincisal distance < 20 mm Aged > 18 years Provision of written informed consent for the study Not having been treated for OSF either medically or surgically Exclusion criteria None specified
Interventions	Intervention characteristics Surgery and placental extract (Group S) Description of intervention: bilateral buccal mucosal fibrotomy, bilateral coronoidectomy, repaired with buccal pad of fat. Postoperatively, placental extract was applied topically every 4 hours for 4 weeks. Follow‐up period: 4 weeks Concurrent treatment provided alongside intervention: physiotherapy and oral hygiene instruction Surgery (Group C) Description of intervention: bilateral buccal mucosal fibrotomy, bilateral coronoidectomy, repaired with buccal pad of fat Follow‐up period: 4 weeks Concurrent treatment provided alongside intervention: physiotherapy and oral hygiene instruction
Outcomes	Interincisal mouth opening: mm Postoperative discomfort: VAS Subjective assessment of wound and wound granulation: scores of 1–6 assigned (1 = no granulation tissue, or presence of necrotic debris; 2 = granulation tissue < 25% of wound; 3 = granulation tissue 25–75% of wound, with presence of epithelisation > 25% of wound; 4 = granulation tissue 75–100% of wound, with presence of epithelisation > 50% of wound; 5 = granulation tissue 100% of wound, with presence of epithelisation > 75% of wound)
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Simple random sampling method."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No placebo used.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Evaluators unaware of allocation.
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No exclusions or attrition.
Selective reporting (reporting bias)	Low risk	All predetermined outcomes reported.
Other bias	Low risk	No concerns.

Vadepally 2019.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Physiotherapy by wooden tongue depressors (Group 1) Initial participant number randomised to group: 25 Physiotherapy by Heister jaw opener (Group 2) Initial participant number randomised to group: 25 Overall Initial participant number: 50 Inclusion criteria Mouth opening < 25 mm (OSF stage III–IVa) Clinical and histological diagnosis OSF Palpable fibrous bands in buccal mucosa Blanched oral mucosa Limited tongue movements Exclusion criteria OSF stage I or II People with malignant transformation People with medical compromise Unable to perform mouth opening exercises on their own Unwilling to cease areca nut chewing habit Unwilling to take part in study Unco‐operative with postoperative physiotherapy regimen
Interventions	Intervention characteristics Physiotherapy by wooden tongue depressors (Group 1) Description of intervention: mouth opening physiotherapy using placement of wooden tongue depressors between teeth 5 or 6 times daily for 6 months, followed by another 12 months at reduced frequency Follow‐up period: 18 months Concurrent treatment provided alongside intervention: fibrous band incision with buccal fat pad flap repair, bilateral coronoidectomy, and extraction of third molars. Oral supplements of vitamin A 50,000 IU, vitamin B complex 200 mg, vitamin C 500 mg, and beta‐carotene, along with topical application of triamcinolone acetonide 0.1% applied over the surgical area for ≥ 6 months postoperatively Physiotherapy by Heister jaw opener (Group 2) Description of intervention: mouth opening physiotherapy using Heister jaw opener instrument between teeth 5 or 6 times daily for 6 months, followed by another 12 months at reduced frequency Follow‐up period: 18 months Concurrent treatment provided alongside intervention: fibrous band incision with buccal fat pad flap repair, bilateral coronoidectomy, and extraction of third molars. Oral supplements of vitamin A 50,000 IU, vitamin B complex 200 mg, vitamin C 500 mg, and beta‐carotene, along with topical application of triamcinolone acetonide 0.1% applied over the surgical area for ≥ 6 months postoperatively
Outcomes	Interincisal mouth opening: mm Burning sensation: VAS Compliance to intervention: unclear Adverse events: presence of adverse events
Identification	Sponsorship source: none Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The patients included in the study were grouped by simple randomization using the envelope method of sampling, resulting in a double blind study."
Allocation concealment (selection bias)	Low risk	Quote: "The patients included in the study were grouped by simple randomization using the envelope method of sampling, resulting in a double blind study."
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding not feasible.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Quote: "Two blinded investigators recorded the mouth opening and ratings of outcomes at regular follow up."
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to make a judgement.
Selective reporting (reporting bias)	High risk	Burning sensation not reported for all time points.
Other bias	Low risk	No concerns.

Wu 2010.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Prednisolone acetate plus salvia miltorrhiza injection – medium‐stage cases (Group 1a) Initial participant number randomised to group: 30 Prednisolone acetate plus salvia miltorrhiza injection – advanced‐stage cases (Group 1b) Initial participant number randomised to group: 30 Prednisolone acetate plus lidocaine injection – medium‐stage cases (Group 2a) Initial participant number randomised to group: 30 Prednisolone acetate plus lidocaine injection – advanced‐stage cases (Group 2b) Initial participant number randomised to group: 30 Overall Initial participant number: 120 Inclusion criteria Medium‐ and advanced‐stage cases Reported irritant pain or appearance of blisters when taking stimulating food Obvious pale palate Palpable stripes on buccal mucosa, pterygomandibular ligament area, soft palate and tongue Mouth opening limitation No history of treatment No history of serious systemic diseases or other oral mucosal diseases No other medication history Exclusion criteria Abnormal hepatorenal function Unable to be treated regularly "Incomplete treatment records which can affect the efficacy judgement"
Interventions	Intervention characteristics Prednisolone acetate plus salvia miltorrhiza injection – medium‐stage cases (Group 1a) Description of intervention: 2.0 mL of prednisolone acetate injection (25 g/L) with 1.0 mL of lidocaine (20 g/L), and 2.0 mL of compound salvia miltiorrhiza injection (1.5 g/mL) with 1.0 mL of lidocaine (20 g/L), given intralesionally once a week for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Prednisolone acetate plus salvia miltorrhiza injection – advanced stage cases (Group 1b) Description of intervention: 2.0 mL of prednisolone acetate injection (25 g/L) with 1.0 mL of lidocaine (20 g/L), and 2.0 mL of compound salvia miltiorrhiza injection (1.5 g/mL) with 1.0 mL of lidocaine (20 g/L), given intralesionally once a week for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Prednisolone acetate plus lidocaine injection – medium‐stage cases (Group 2a) Description of intervention: 2.0 mL of prednisolone acetate injection (25 g/L) with 1.0 mL of lidocaine (20 g/L), given intralesionally once a week for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none Prednisolone acetate plus lidocaine injection – advanced‐stage cases (Group 2b) Description of intervention: 2.0 mL of prednisolone acetate injection (25 g/L) with 1.0 mL of lidocaine (20 g/L), given intralesionally once a week for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: none
Outcomes	Interincisal mouth opening: cm Size of grey lesion area: square cm Therapeutic efficacy rate: percentage Adverse effect rate: percentage
Identification	Sponsorship source: Education Development Fund of Hunan, P.R. China Country: China
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to make judgement.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded – groups treated differently.
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low attrition.
Selective reporting (reporting bias)	Low risk	All predetermined outcomes reported.
Other bias	Low risk	No concerns.

Yadav 2014.

Study characteristics
Methods	Study design: parallel‐group RCT
Participants	Baseline characteristics Dexamethasone (Group 1) Initial participant number randomised to group: 20 Turmix (Group 2) Initial participant number randomised to group: 20 Overall Initial participant number: 40 Inclusion criteria Aged 20–40 years Histological diagnosis of OSF Exclusion criteria Co‐existing malignancy or white lesions Known allergies to drugs Previously treated for OSF Not willing to have biopsy
Interventions	Intervention characteristics Dexamethasone (Group 1) Description of intervention: intralesional injection of dexamethasone (4 mg/mL), hyaluronidase 1500 IU and 0.5 mL of 2% lignocaine with adrenaline, administered weekly at multiple sites, alternating sides, for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis Turmix (Group 2) Description of intervention: 2 tablets of Turmix (Curcumin longa 300 mg, Piperine 5 mg) orally once daily for 3 months Follow‐up period: 3 months Concurrent treatment provided alongside intervention: habit cessation advice and oral prophylaxis
Outcomes	Burning sensation improvement: percentage reduction of VAS score Interincisal mouth opening: mm Tongue protrusion: mm
Identification	Sponsorship source: not declared Country: India
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The subjects with confirmed histopathological diagnosis were then randomly divided into 2 groups of 20 each by using table of random numbers."
Allocation concealment (selection bias)	Unclear risk	Insufficient information to make judgement
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open label
Blinding of outcome assessment (detection bias) Mouth opening	Low risk	Lack of assessor blinding unlikely to affect outcome
Blinding of outcome assessment (detection bias) Burning sensation	High risk	Participants not blinded
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Final group populations not reported
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	No concerns

ASA: American Society of Anesthesiologists; cm: centimetre; Ig: immunoglobulin; mm: millimetre; OSF: oral submucous fibrosis; RCT: randomised controlled trial; TMJ: temporomandibular joint; VAS: visual analogue scale.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Agarwal 2011	Not an RCT
Agrawal 2018	Not an RCT
Ambereen 2021	No histological diagnosis
Angadi 2011	Not an RCT
Anuradha 2017	No histological diagnosis
Ara 2016	Not an RCT
Awan 2014	Not an RCT
Aziz 2009	Not an RCT
Bande 2013	Not an RCT
Bande 2016	Not an RCT
Baptist 2016	No histological diagnosis
Biswas 2015	No histological diagnosis
Bohra 2021	Not an RCT
Borle 2009	Not an RCT
Chandrashekar 2021	No histological diagnosis
Chen 2021	No histological diagnosis
Chole 2012	Not an RCT
CTRI/2021/09/036499	No histopathological diagnosis
Daga 2017	No histological diagnosis
Dasukil 2022	No histological diagnosis
Datarkar 2020	No histological diagnosis
Deepak 2021	No histological diagnosis
Ganguly 2020	No histological diagnosis
Gupta 1988	Not an RCT
Gupta 2022	No histopathological diagnosis
Hazarey 2015	No histological diagnosis
Idrees 2016	Not an RCT
Jain 2000	Full text not available
Jiang 2009	Not an RCT
Jiang 2015	No histological diagnosis
Johny 2019	Not an RCT
Kakar 1985	Not an RCT
Kamath 2014	Not an RCT
Kanjani 2019	No histological diagnosis
Karemore 2012	Not an RCT
Kerr 2011	Not an RCT
Kisave 2020	No histological diagnosis
Krishnamoorthy 2013	Does not include primary or secondary outcomes specified for this review
Lambade 2015	Not an RCT
Lanjekar 2020	No histological diagnosis
Li 2018	Unable to isolate participants with OSF from participants with oral cancer
Liu 2017	Not an RCT
Lu 2019	Not an RCT
Mathai 2017	Does not include primary or secondary outcomes specified for this review
Mehrotra 2009	Not an RCT
Memon 2022	No histopathological diagnosis
Mostafa 2021	No histological diagnosis
Nihadha 2022	No histopathological diagnosis
Pandey 2020	No histopathological diagnosis
Pardeshi 2015	Not an RCT
Patil 2016a	No histological diagnosis
Patil 2016b	No histological diagnosis
Pentapati 2016	Not an RCT
Pentapati 2017	Not an RCT
Pipalia 2016	No histological diagnosis
Pipalia 2017	Not an RCT
Rai 2014	Not an RCT
Rai 2019	No histological diagnosis
Raizada 2017	No histological diagnosis
Raizada 2022	No histopathological diagnosis
Rizvi 2019	No histological diagnosis
Saalim 2020	No histological diagnosis
Sadaksharam 2017	No histological diagnosis
Saran 2018	No histological diagnosis
Selvam 2013	No histological diagnosis
Shams 2022	No histopathological diagnosis
Shandilya 2021	No histological diagnosis
Sharma 2012	Not an RCT
Shetty 2013	Not an RCT
Singh 2015	Not an RCT
Singh 2016a	No histological diagnosis
Singh 2016b	Not an RCT
Singh 2018	No histological diagnosis
Srivastava 2021	No histological diagnosis
Tai 2001	Not an RCT
Tp 2019	No histological diagnosis
Vibha 2019	Not an RCT
Warnakulasuriya 2016	Not an RCT
Wollina 2015	Not an RCT

OSF: oral submucous fibrosis; RCT: randomised controlled trial.

Characteristics of studies awaiting classification [ordered by study ID]

Alam 2013.

Methods	Study design: 4‐arm parallel‐group RCT
Participants	Baseline characteristics Medicinal treatment plus aloe vera gel (Group A Med) Initial participant number randomised to group: 15 Medicinal treatment plus control (Group B Med) Initial participant number randomised to group: 15 Surgical treatment plus aloe vera gel (Group B Surg) Initial participant number randomised to group: 15 Surgical treatment plus control (Group B Surg) Initial participant number randomised to group: 15 Overall Initial participant number: 60 (53 male, 7 female) Inclusion criteria: 'diagnosed' OSF (detail of diagnostic criteria not reported) Exclusion criteria: uncontrolled diabetes, compromised immunity, chronic infection
Interventions	Intervention characteristics Medicinal treatment plus aloe vera gel Description of intervention: submucosal injections twice a week of hyaluronidase 1500 IU diluted with 1 mL of 2% lignocaine (with 1:80,000 adrenaline) for the first 3 weeks. This was followed by submucosal injection twice a week of hyaluronidase 1500 IU diluted in dexamethasone 4 mg and 1 mL of 2% lignocaine (with 1:80,000 adrenaline) for the next 7 weeks. Massaging the cheek with the mouth closed, followed by physiotherapy with a Heister mouth gag for 20 minutes, performed after the submucosal injection Aloe vera: Group A participants were advised to apply aloe vera gel over the buccal mucosa, palate, retromolar region and floor of the mouth twice daily during the treatment as well as in the follow‐up phase, up to 6 months after the completion of medicinal or surgical treatment Duration of intervention: 6 months Follow‐up period: 6 months Concurrent treatment provided alongside intervention: participants were actively discouraged from consuming the identified aetiological factors, such as pan masala, gutkha, betel quid, tobacco, and other chronic irritants such as hot and spicy food. Participants with anaemia were treated and encouraged to eat a well‐balanced diet. All received a supplementary therapeutic regimen of Lycostar (lycopene 5000 mg; carotene 10.33 mg; refined wheat germ oil 25 mg; zinc sulphate 27.45 mg and selenium dioxide 75 mg; Mankind Pharma Ltd, New Delhi, India) twice daily and Capsule Becosules‐Z (thiamine 10 mg; riboflavin 10 mg; pyridoxine, 3 mg; vitamin B12 15 mg; niacinamide 100 mg; calcium pantothenate 50 mg; folic acid 1.5 mg; biotin 100 mg; ascorbic acid 50 mg and zinc sulphate 41.4 mg, equivalent to of elemental zinc 15 mg; Pfizer Ltd, USA) once daily during the treatment and up to 6 months after the completion of treatment. They were advised to perform physiotherapy for mouth opening 4 or 5 times a day on a regular basis. Impacted or malposed third molars having possible correlation with the prognosis of OSF treatment were removed. All possible foci of infections were eradicated. Correction of local irritants, such as a sharp tooth, was performed. Fractured and carious teeth were restored. Oral prophylaxis was provided. Participants were assessed after 2 weeks; and after they gave assurance that they had suspended the habit as instructed, further management was continued. Description of adverse events: none reported Medicinal treatment plus control (Group Med B) Description of intervention: submucosal injections twice a week of hyaluronidase 1500 IU diluted with 1 mL of 2% lignocaine (with 1:80,000 adrenaline) for the first 3 weeks. This was followed by submucosal injection twice a week of hyaluronidase 1500 IU diluted in dexamethasone 4 mg and 1 mL of 2% lignocaine (with 1:80,000 adrenaline) for the next 7 weeks. Massaging the cheek with the mouth closed, followed by physiotherapy with a Heister mouth gag for 20 minutes, was done after the submucosal injection Aloe vera: no aloe vera gel Duration of intervention: 7 weeks plus 6 months physiotherapy Follow‐up period: 6 months Concurrent treatment provided alongside intervention: participants were actively discouraged from consuming the identified aetiological factors, such as pan masala, gutkha, betel quid, tobacco, and other chronic irritants such as hot and spicy food. Participants with anaemia were treated and encouraged to eat a well‐balanced diet. All participants were put on a supplementary therapeutic regimen of Lycostar (lycopene 5000 mg; carotene 10.33 mg; refined wheat germ oil 25 mg; zinc sulphate 27.45 mg and selenium dioxide 75 mg; MankindPharma Ltd, New Delhi, India) twice daily and Capsule Becosules‐Z (thiamine 10 mg; riboflavin 10 mg; pyridoxine 3 mg; vitamin B12 15 mg; niacinamide 100 mg; calcium pantothenate 50 mg; folic acid 1.5 mg; biotin 100 mg; ascorbic acid 50 mg and zinc sulphate 41.4 mg, equivalent to of elemental zinc 15 mg; Pfizer Ltd, USA) once daily during the treatment and up to 6 months after the completion of treatment. They were advised to perform physiotherapy for mouth opening 4 or 5 times a day on a regular basis. Impacted or malposed third molars having possible correlation with the prognosis of OSF treatment were removed. All possible foci of infections were eradicated. Correction of local irritants, such as a sharp tooth, was performed. Fractured and carious teeth were restored. Oral prophylaxis was provided. Participants were assessed after 2 weeks; and after they gave assurance that they had suspended the habit as instructed, further management was continued. Description of adverse events: none reported Surgical treatment plus aloe vera gel Description of intervention: under local or general anaesthesia, the fibrous bands in the buccal mucosa were palpated and incised along the occlusal line, starting from the angle of the mouth extending posteriorly up to the retromolar region, bilaterally, to the depth of connective tissue. A Heister mouthgag was then applied to achieve a maximal interincisal opening of 35 mm, and coronoidectomy was done where required to get the maximal possible opening. Suitable graft (buccal fat pad, nasolabial flap or collagen membrane) was then placed over the mucosal defect. Participants were placed on physiotherapy with a wooden spatula from the fifth postoperative day, ≥ 4 or 5 times a day, for a minimum of 6 months. Aloe vera: Group A participants were advised to apply aloe vera gel over the buccal mucosa, palate, retromolar region, and floor of the mouth, twice daily during the treatment, as well as in the follow‐up phase, up to 6 months after the completion of medicinal or surgical treatment Duration of intervention: one‐off intervention plus 6 months' home physiotherapy Follow‐up period: 6 months Concurrent treatment provided alongside intervention: participants were actively discouraged from consuming the identified aetiologic factors, such as pan masala, gutkha, betel quid, tobacco, and other chronic irritants such as hot and spicy food. Participants with anaemia were treated and encouraged to eat a well‐balanced diet. All participants were put on a supplementary therapeutic regimen of Lycostar (lycopene 5000 mg; carotene 10.33 mg; refined wheat germ oil 25 mg; zinc sulphate 27.45 mg and selenium dioxide 75 mg; Mankind Pharma Ltd, New Delhi, India) twice daily and Capsule Becosules‐Z (thiamine 10 mg; riboflavin 10 mg; pyridoxine 3 mg; vitamin B12 15 mg; niacinamide 100 mg; calcium pantothenate 50 mg; folic acid 1.5 mg; biotin 100 mg; ascorbic acid 50 mg and zinc sulphate 41.4 mg, equivalent to of elemental zinc 15 mg; Pfizer Ltd, USA) once daily during the treatment and up to 6 months after the completion of treatment. They were advised to perform physiotherapy for mouth opening 4 or 5 times a day on a regular basis. Impacted or malposed third molars having possible correlation with the prognosis of OSF treatment. All possible foci of infections were eradicated. Correction of local irritants, such as a sharp tooth, was performed. Fractured and carious teeth were restored. Oral prophylaxis was provided. Participants were assessed after 2 weeks; and after they gave assurance that they had suspended the habit as instructed, further management was continued. Description of adverse events: none reported Surgical treatment plus control (Group Surg B) Description of intervention: under local or general anaesthesia required in the individual case, the fibrous bands in the buccal mucosa were palpated and incised along the occlusal line, starting from the angle of the mouth extending posteriorly up to the retromolar region, bilaterally, to the depth of connective tissue. A Heister mouthgag was then applied to achieve a maximal interincisal opening of 35 mm, and coronoidectomy was done where required to get the maximal possible opening. Suitable graft (buccal fat pad, nasolabial flap or collagen membrane) was then placed over the mucosal defect. Participants were placed on physiotherapy with a wooden spatula from the fifth postoperative day, ≥ 4 or 5 times a day, for a minimum of 6 months. Aloe vera: no aloe vera gel Duration of intervention: one‐off intervention plus 6 months home physiotherapy Follow‐up period: 6 months Concurrent treatment provided alongside intervention: participants were actively discouraged from consuming the identified aetiologic factors, such as pan masala, gutkha, betel quid, tobacco, and other chronic irritants such as hot and spicy food. Participants with anaemia were treated and encouraged to eat a well‐balanced diet. All participants were put on a supplementary therapeutic regimen of Lycostar (lycopene 5000 mg; carotene 10.33 mg; refined wheat germ oil 25 mg; zinc sulphate 27.45 mg and selenium dioxide 75 mg; MankindPharma Ltd, New Delhi, India) twice daily and Capsule Becosules‐Z (thiamine 10 mg; riboflavin 10 mg; pyridoxine 3 mg; vitamin B12 15 mg; niacinamide 100 mg; calcium pantothenate 50 mg; folic acid 1.5 mg; biotin 100 mg; ascorbic acid 50 mg and zinc sulphate 41.4 mg, equivalent to of elemental zinc 15 mg; Pfizer Ltd, USA) once daily during the treatment and up to 6 months after the completion of treatment. They were advised to perform physiotherapy for mouth opening 4 or 5 times a day on a regular basis. Impacted or malposed third molars having possible correlation with the prognosis of OSF treatment were removed. All possible foci of infections were eradicated. Correction of local irritants, such as a sharp tooth, was performed. Fractured and carious teeth were restored. Oral prophylaxis was provided. Participants were assessed after 2 weeks; and after they gave assurance that they had suspended the habit as instructed, further management was continued. Description of adverse events: Not reported
Outcomes	Interincisal mouth opening Outcome type: continuous Reporting: fully reported Unit of measure: mm Direction: higher is better Burning sensation Outcome type: continuous Reporting: partially reported Scale: VAS Range: 0–10 Direction: lower is better Tongue protrusion Outcome type: continuous Reporting: fully reported Unit of measure: mm Direction: higher is better Inter‐earlobe distance on cheek blowing Outcome type: continuous Reporting: fully reported Unit of measure: mm Compliance to intervention Outcome type: dichotomous Reporting: fully reported Adverse events Outcome type: adverse event Reporting: fully reported Burning sensation before and after (paper reported figures) Outcome type: continuous Reporting: fully reported Scale: VAS Mouth opening before and after (paper reported figures) Outcome type: continuous Tongue protrusion before and after (paper reported figures) Outcome type: continuous Earlobe before and after (paper reported figures) Outcome type: continuous
Notes	Incomplete data reporting – awaiting further information from authors

CTRI/2022/03/041108.

Methods	Parallel‐group RCT with stratified block randomisation; allocation by sequentially numbered, sealed, opaque envelopes; outcome assessor blinding
Participants	Inclusion criteria Consecutive patients with clinically suspected OSF will be considered for the study. People with current or previous use of any form of smokeless tobacco or areca nut (or both) (any duration and frequency) With late‐stage OSF designated as Grade III and IV by the grading system proposed by "Kerr et al" Willing to quit the habit of smokeless tobacco or areca nut, poor both Willing to accept regular follow‐up protocol Histopathological confirmation of OSF by incisional biopsy Exclusion criteria People with early‐stage OSF designated as Grade I and II by the grading system proposed by "Kerr et al" With known allergy or contraindication to COX‐2 inhibitors, NSAIDs, salicylates or sulphonamides With diagnosis or treatment of oesophageal, gastric, pyloric channel or duodenal ulceration within past 30 days, prior or active pancreatic disease or inflammatory bowel disease or haemophilia or a history of bleeding disorders or gout or a history of gout Already undergoing treatment for OSF Pregnant and lactating females With concomitant pericoronitis or any other dental condition leading to trismus
Interventions	Intervention: oral COX‐2 inhibitors, surgery and aspirin mouth rinse: celecoxib 100 mg tablet twice daily orally for 12 weeks followed by fibrotomy and buccal pad of fat graft, followed by aspirin mouth rinse twice a day for 12 months along with physiotherapy Control intervention: antioxidants followed by surgery followed by placebo mouth rinse: cap antioxidants twice daily orally for 12 weeks, followed by fibrotomy and buccal pad of fat graft, followed by placebo mouth rinse twice a day for 12 months
Outcomes	Primary outcomes Clinical response measured by: interincisal mouth opening tongue protrusion burning sensation/pain relapse clinical signs of malignant transformation Participants will be followed up every 2 weeks during the 12‐week presurgical period, and on the 7th, 15th, 30th day, and of the 3rd, 6th, 9th, and 12th month postsurgery.
Notes

Wu 2015.

Methods
Participants
Interventions
Outcomes
Notes	Awaiting further information from authors

COX: cyclo‐oxygenase; NSAID: non‐steroidal anti‐inflammatory drug; OSF: oral submucous fibrosis; RCT: randomised controlled trial.

Characteristics of ongoing studies [ordered by study ID]

CTRI201202002437.

Study name	Comparison of steroid releasing patch and steroid injections for oral submucous fibrosis
Methods	Phase 2 parallel RCT
Participants	Target sample size: 24 Inclusion criteria: any stage of oral submucous fibrosis, willing to stop the habit of chewing betel quid/Gutkha, no systemic disorders such as diabetes mellitus, AIDS or hypertension Exclusion criteria: not stopping chewing habit, not willing to get intralesional injections, unable to come repeatedly for treatment/assessment
Interventions	Betamethasone injections administered at 3 different sites of buccal mucosa bilaterally using insulin syringe twice per week (once in 3 days, excluding Sunday). The sites of injection were centre of buccal mucosa in 3rd molar, 1st molar and 1st premolar region.
Outcomes	Primary outcomes Mouth opening measured as interincisal distance in mm Burning sensation measured as 0–100 VAS Time point: every week after treatment Secondary outcomes Discomfort level during the treatment procedure recorded in participant's own words as descriptive data Time point: once in a week after treatment received
Starting date	First enrolment: 15 July 2009
Contact information	drujjwaljoshi@gmail.com
Notes

CTRI201305003609.

Study name	Exercise device to increase mouth opening in reduced mouth opening conditions
Methods	Parallel RCT, outcome assessor blinded
Participants	Inclusion criteria: people aged 18–35 years with OSF who agree to participate in the clinical trial and assure co‐operation with the intervention and follow‐up visits Exclusion criteria: history of systemic disease with hepatic, respiratory, renal, cardiac or haematological disease; presence of any malignancy associated; persistent infection or severe infection in the last 3 months; cannot be relied upon to comply or unwilling to sign informed consent; requiring treatment during the study period with drugs not permitted by the study protocol
Interventions	Intervention: mouth exercising device group 1. Oral antioxidant: once in a day + local steroid ointments twice a day and ice‐cream‐stick regimen for mouth exercise twice a day + mouth exercising device 2–3 times daily for 10 minutes on each side. 2. (Point 1) + additional intralesional steroid injection. 1 mL steroid in a local anaesthetic agent to be injected intralesionally + mouth exercising device 2–3 times daily for 10 minutes on each side. 3. (Point 1) + surgery (for mouth opening < 20 mm) + mouth exercising device 2–3 times daily for 10 minutes on each side after surgical mouth opening Duration of therapy: 6 months Control 1. Oral antioxidant: once in a day + local steroid ointments twice a day and ice‐cream‐stick regimen for mouth exercise twice a day. 2. (Point 1) + additional intralesional steroid injection. 1 mL steroid in a local anaesthetic agent to be injected intralesionally. 3. (Point 1) + surgery (for mouth opening < 20 mm) Duration of therapy: 6 months
Outcomes	Primary outcomes Interincisal distance on maximum mouth opening in mm Burning sensation – VAS in percentage Secondary outcomes Difficulty in eating and swallowing Timepoint: 6 months
Starting date	First enrolment: 18 July 2011
Contact information	pravinandsmita@yahoo.co.in
Notes

CTRI201504005696.

Study name	A randomised control trial of the corticosteroids in OSF
Methods	Parallel RCT
Participants	60
Interventions	Intervention: injection betamethasone and injection hyaluronidase: injections of betamethasone sodium phosphate, 4 mg/mL (Betnesol‐GSK GlaxoSmithKline Pharmaceuticals Limited, Dr Annie Beasant Road, Worli, Mumbai, India) and hyaluronidase 1500 IU/mL (Hylase‐Shreya Life Sciences Private Limited, Periear Hill Road, Andheri, Mumbai, India) Control: placebo – saline injection, weekly injections
Outcomes	Primary outcome Mouth opening Secondary outcome Burning sensation in the oral cavity Timepoint: weekly for 10 weeks
Starting date	26 June 2014
Contact information	Kaustubh Sansare, Nair Hospital Dental College, A.L.Nair Road, Mumbai Central, Mumbai 400008, Maharashtra, India Email: kaustubhsansare@yahoo.com
Notes

CTRI201701007732.

Study name	Effect of two drugs for treatment of oral submucous fibrosis
Methods	Parallel RCT
Participants	56 Inclusion criteria: age 20–60 years with clinically and histopathologically confirmed case of OSF; a positive history of chewing areca nut or 1 of its commercial preparations, difficulty in swallowing and chewing, and burning sensation on eating spicy foods; restricted mouth opening and changes in the oral mucous membrane including the presence of palpable vertical fibrous bands, stiffness and blanching; positive habit history of tobacco, smoking, betel quid, and alcohol Exclusion criteria: not willing to participate; already undertaking some treatment for existing potentially malignant oral disorder; severe systemic disease
Interventions	Intervention: tablet curcumin and piperine 300 mg tablet administered orally 3 times daily for 9 months (Group A) Comparator: tablet pentoxifylline 400 mg tablet administered orally 3 times daily for 9 months (Group B)
Outcomes	Evaluation of burning sensation score, mouth opening, tongue protrusion and cheek flexibility between groups and within group Participants were assessed clinically every 4 weeks for 9 months
Starting date	1 October 2012
Contact information	Dr Shikha, Department of Oral Medicine & Radiology, Maulana Azad Institute of Dental Sciences, New Delhi, Delhi, 110002, India calmshikhs@yahoo.com
Notes

CTRI201709009666.

Study name	Comparison of aloe vera gel and curcumin gel in oral submucous fibrosis
Methods	Parallel RCT
Participants	105 Inclusion criteria: age 18–60 years; newly diagnosed with OFS; older people with submucous fibrosis who have left treatment 6 weeks before Exclusion criteria: chronic diseases such as tuberculosis, diabetes mellitus, AIDS; any other oral mucosal lesions along with oral submucous fibrosis; pregnant and lactating women
Interventions	Intervention: aloe vera gel with intralesional triamcinolone 40 mg, topical application, 3 times a day for 12 weeks Intervention: curcumin gel with intralesional triamcinolone 40 mg, topical application 3 times a day for 12 weeks Comparator: topical vehicle (polyethylene glycol) with intralesional triamcinolone 40 mg, topical application 3 times a day for 12 weeks
Outcomes	Improvement in burning sensation Interincisal distance Tongue protusion All evaluated at the end of 4, 8 and 12 weeks
Starting date	8 March 2017
Contact information	Darakhshan Rizvi, Eras Lucknow Medical College & Hospital, Sarfarazganj Hardoi Road, Lucknow, Uttar Pradesh, 226003, India rizvidarakhshan@gmail.com
Notes	Unclear how OSF will be diagnosed

CTRI201709009674.

Study name	Effect of two drugs in treatment of oral submucous fibrosis
Methods	Parallel RCT
Participants	45 Inclusion criteria: age 20–70 years with clinically and histopathologically confirmed OSF; positive history of chewing areca nut, tobacco, betel quid, smoking and alcohol; difficulty in swallowing, chewing and burning sensation on eating spicy foods; restricted mouth opening and changes in the oral mucous membrane including the presence of palpable vertical fibrous bands, stiffness and blanching Exclusion criteria: not willing to participate; already undertaking some treatment for existing potentially malignant oral disorder; severe systemic disorder
Interventions	Intervention: tablet curcumin and piperine 300 mg administered orally twice daily for 6 months (Group A) Intervention: tablet lycopene 16 mg administered orally twice daily for 6 months (Group B) Comparator: placebo tablets, manufactured by Sanat Pharmaceuticals dispensed in bottle with label
Outcomes	Evaluation of burning sensation score, mouth opening, tongue protrusion and cheek flexibility between groups and within group clinically every 4 weeks for 9 months Reporting of malignant transformation over period of 9 months; participants were assessed clinically every 4 weeks for 9 months. In cases of any suspicion, biopsy was performed.
Starting date	1 October 2015
Contact information	Dr Anshul Mahajan, Department of Oral Medicine and Radiology Maulana Azad Institute of Dental Sciences New Delhi, Central, Delhi, 110002, India anshulmahajan@hotmail.com
Notes

CTRI201804013147.

Study name	Comparison of different treatment options in oral lichen planus and oral submucous fibrosis
Methods	Parallel RCT
Participants	120 Inclusion criteria: age 15–70 years with clinically and histopathologically confirmed cases of oral lichen planus Exclusion criteria: unwilling to participate; undergoing alternative treatment for the same condition; severe uncontrolled systemic disease and pregnancy
Interventions	Intervention: topical aloe vera gel (for oral lichen planus). Topical application of aloe vera extract 500 mg capsules 3 times daily for 8 weeks Comparator: low‐level laser therapy (for oral lichen planus) using a 980 nm diode laser at 0.3 W and 30 seconds per site using a non‐contact mode twice weekly for 8 weeks Intervention: topical aloe vera gel (for oral submucous fibrosis). Topical application of aloe vera extract 500 mg capsules 3 times daily for 8 weeks Comparator: intralesional hyaluronidase and dexamethasone (for oral submucous fibrosis), intralesional hyaluronidase injection 1500 IU with 5 mL dexamethasone once weekly for 8 weeks
Outcomes	Mouth opening, tongue protrusion, cheek flexibility and VAS score at the time of first visit and every 4 weeks for 9 months Reporting of any malignant transformation over a period of 9 months – individuals will be assessed every 4 weeks, and in cases of any suspicion, biopsy will be done.
Starting date	12 April 2018
Contact information	Dr Sunita Gupta, Department of Oral Medicine and Radiology, Maulana Azad Institute of Dental Sciences, BSZ Marg, New Delhi, Central, Delhi, 110002, India Email: drsunitagupta@yahoo.com
Notes	OLP data published but no OSF data published

CTRI201804013285.

Study name	Efficacy of antioxidants in the management of oral submucous fibrosis
Methods	Parallel RCT
Participants	50 Inclusion criteria: age 18–80 years Exclusion criteria: hypersensitivity to curcumin or lycopene (or both); gallstones and other gastrointestinal disorders; fibrotic lesions other than oral submucous fibrosis; receiving any other medications
Interventions	Intervention: lycopene capsules 2000 μg twice daily orally for 3 months Comparator: combined therapy with lycopene and curcumin 2000 μg lycopene twice daily and circuit h 400 mg once daily orally for 3 months
Outcomes	Relief from burning sensation, improved interincisal mouth opening, cheek mucosa flexibility improvement Timepoints: end of first month and at the end of 3 months after starting the therapy
Starting date	1 November 2017
Contact information	Sudeepthi Pulimi, Department of Oral Medicine and Radiology, Narayana Dental College, Nellore, Andfra Pradesh, 524003 India Email: dr.pulim@gmail.com
Notes

min: minutes; mm: millimetres; nm: nanometres; OLP: oral lichen planus; OSF: oral mucous fibrosis; RCT: randomised controlled trial; sec: seconds; VAS: visual analogue scale; W: watts.

Differences between protocol and review

The primary outcomes have been resequenced from the order they appear in the published protocol.

• Resumption of normal eating, chewing and speech

• Change or improvement in maximal jaw opening, measured as the interincisal distance

• Improvement in range of jaw movement utilising any validated assessment tool

• Change in severity of oral/mucosal burning pain using any recognised validated pain scale

Contributions of authors

AJ designed the updated review protocol, screened abstracts and full‐text studies, carried out risk of bias assessments, extracted and analysed data, carried out GRADE evidence certainty assessments, and drafted and revised the full report.

BV designed the updated review protocol, screened abstracts and full‐text studies, carried out risk of bias assessments, extracted data, and commented on and revised the full report.

TL designed the updated review protocol, screened abstracts and full‐text studies, carried out risk of bias assessments, extracted data, and drafted and revised the full report.

VRA screened full‐text studies, and commented on and revised the full report.

JT screened full‐text studies, analysed data, carried out GRADE evidence certainty assessments, and drafted and revised the full report.

Sources of support

Internal sources

• Division of Dentistry, School of Medical Sciences, Faculty of Biology, Medicine and Health, The University of Manchester; Manchester Academic Health Sciences Centre (MAHSC); and the NIHR Manchester Biomedical Research Centre, UK

  Support from Cochrane Oral Health at The University of Manchester

External sources

• National Institute for Health Research (NIHR), UK

  CRG funding acknowledgement:
The NIHR was the largest single funder of the Cochrane Oral Health Group until March 2023.
Disclaimer:
The views and opinions expressed herein are those of the authors and do not necessarily reflect those of the NIHR, National Health Service or the Department of Health.

• Cochrane Oral Health Global Alliance, Other

  Cochrane Oral Health reviews have previously been supported by Global Alliance member organisations (British Association of Oral Surgeons, UK; British Orthodontic Society, UK; British Society of Paediatric Dentistry, UK; British Society of Periodontology, UK; Canadian Dental Hygienists Association, Canada; National Center for Dental Hygiene Research & Practice, USA; Mayo Clinic, USA; New York University College of Dentistry, USA; and Royal College of Surgeons of Edinburgh, UK) providing funding for the editorial process (prior to March 2023).

• The University of Pennsylvania, USA

  Funding acknowledgement:
The work of Cochrane Oral Health (COH) is supported by a collaborative research agreement between The University of Manchester and The University of Pennsylvania. The research collaboration sees the creation of a Cochrane Oral Health Collaborating Center at the University of Pennsylvania School of Dental Medicine, Center for Integrative Global Oral Health, which will work alongside COH (Manchester).
Disclaimer:
The views and opinions expressed herein are those of the authors and do not necessarily reflect those of either The University of Pennsylvania or The University of Manchester.

Declarations of interest

AJ: none.

BV: none.

TL: none.

VRA: none.

JT: none.

New search for studies and content updated (conclusions changed)