Introduction

Mucosal ulceration is characterized by tissue loss, which can affect both the epithelial and underlying connective tissues. Oral ulcers typically result in variable painful symptoms depending on the location and individual characteristics [1,2,3]. Ulcerative lesions are common manifestations of several pathologies with diverse etiological factors and may also be associated with systemic diseases [3, 4].

Ulcers are classified based on their duration (with lesions lasting longer than two weeks considered chronic), number (single or multiple), and etiology (infectious, traumatic, autoimmune, neoplastic). Accurate diagnosis requires a thorough patient history and detailed physical examination, often complemented by additional tests [4, 5]. The main causes of oral ulcerative lesions are as follows: trauma (chemical or physical); infections; autoimmune conditions such as recurrent aphthous ulceration, oral lichen planus (OLP), pemphigus vulgaris, and lupus erythematosus; and as a consequence of antineoplastic therapy - oral mucositis resulting from chemotherapy or radiotherapy in the head and neck region [3,4,5,6,7].

Oral inflammatory ulcers are typically painful and can cause a significant decrease in patients’ quality of life by negatively impacting basic functions such as speaking and eating [8]. Therefore, even in cases of self-limiting ulcers, therapeutic strategies may be necessary. Treatment may vary according to the etiological factor [9]. To resolve traumatic ulcers, it is necessary to remove the factor causing the injury (misadapted prostheses, sharp edges of teeth, etc.), and often analgesics may be used for pain control. For infectious ulcers, treatment depends on the causative microorganism, and antibiotics or antivirals may be used [10]. In such cases, the use of low-power lasers with the antimicrobial photodynamic therapy (aPDT) protocol can also contribute to ulcer disinfection [11, 12]. In cases of immune-mediated lesions, topical corticosteroids such as clobetasol propionate can be employed. However, depending on the severity, systemic corticosteroids or other immunomodulators may be necessary [8, 9]. In addition to these treatments, ozone therapy has been cited as an alternative method to accelerate wound healing [13].

Ozone therapy has been widely used as an adjunctive treatment for various diseases [14]. In dentistry, it has been applied to manage pain by neutralizing inflammatory mediators, promoting wound healing through enhanced cell proliferation, and reducing contamination of lesions by disrupting the cell membrane of pathogens. Ozone (O3) can be applied to the wound site by incorporating it into water or oil, such as olive oil, or by directly applying gaseous ozone to the tissues [15,16,17,18].

While ozone therapy finds applications in various medical fields, its use in dentistry is still considered a relatively recent development. There is a need for studies to explore the application protocols and examine the effects of ozone therapy on oral tissue healing. Therefore, the objective of this study is to conduct a systematic review and meta-analysis to assess the effectiveness of ozone therapy in healing mouth ulcers, comparing it to placebo or active treatments.

Methods

This systematic review was carried out according to the PRISMA [19] guidelines and it was registered in the PROSPERO system under the protocol number CRD42022307331.

Search strategy

A systematic search was performed in the following databases: Pubmed, EMBASE, Scopus, and Lilacs. The search was performed on 06/08/2023 and repeated on 04/08/2024. The search was based on the following keywords:

(ozone[MeSH Terms] OR ozone therapy OR ozone treatment OR ozonated oil OR ozonotherapy) AND ((mouth diseases[MeSH Terms] OR oral disease OR oral diseases) OR (lichen planus, oral[MeSH Terms] OR oral lichen planus OR OLP) OR (stomatitis[MeSH Terms] OR stomatitis) OR (pemphigus[MeSH Terms] OR oral pemphigus OR pemphigus vulgaris) OR (mucositis[MeSH Terms] OR oral mucositis OR mouth mucositis) OR (oral ulcer[MeSH Terms] OR mouth ulcer OR mouth ulcers OR oral ulceration)). During the research, only studies published in Portuguese, Spanish or English were selected. No other restrictions/limits were applied.

Data collection and analysis

In the first stage of literature selection, titles and abstracts were evaluated by two independent reviewers to identify articles that reported clinical protocols involving ozone therapy for healing of oral ulcers from different etiology. In the second stage, the inclusion and exclusion criteria were applied.

The following inclusion criteria were considered concerning the type of study: clinical trials, prospective and retrospective studies assessing human participants diagnosed with oral inflammatory ulcers of any sex and age. Articles or abstracts with case series including less than 5 subjects were excluded. The inclusion criteria were based on the components of the research question: Population (P), Intervention (I), Comparator (C), and Outcomes (O). P - patients, of any gender and age, diagnosed with oral ulcers. I - oral ulcers treated with ozone therapy. C - placebo or active treatment (corticosteroid and/or laser photobiomodulation). Outcomes - healing of oral ulcer.

Study selection and data extraction and management

After applying the inclusion and exclusion criteria, the selected studies underwent independent methodological evaluation and data extraction by two reviewers.

For each included study, qualitative and quantitative data were extracted: year of publication, number of patients evaluated, study design and follow-up time, intervention and comparison, method of evaluation of results, type of study and any other information necessary for qualitative methodological evaluation. Disagreements were discussed between the two reviewers with a third reviewer until a consensus was reached.

Outcome assessment and meta-analysis

For quantitative analysis and meta-analysis, the primary outcome assessed was clinical response with area reduction of the lesion and symptoms resolution. Secondarily, ulcer duration, Thongprasom score (in case of OLP lesions) and analgesic consumption were also assessed (Table 1).

Table 1 Studies evaluating the ozone therapy in oral ulcers
Full size table

The clinical response was determined by a clinical reduction of the lesion size, while the ulcer duration was determined by the number of days needed for the ulcer to disappear (wound area = 0). The Thongprasom score was based on clinical presentation of OLP lesions observed at baseline, after treatment and follow-up. The analgesic consumption was determined by the number of tablets the participants had taken after the intervention, and the symptom resolution was based on variations in the VAS (visual analog scale).

Statistical analysis was performed using the Review Manager software, version 5.4.1 (Cochrane). The Risk Ratio (RR) and the standardized mean difference (SMD) with 95% CI (confidence interval) or Mean difference (MD) were calculated from continuous data including mean and standard deviation of the results of each study. The heterogeneity of the assays was assessed by the Tau2, Chi2 and I2 statistic test. High heterogeneity was considered when I2 75% [20]. To identify the cause of high heterogeneity, one study in each turn was eliminated to determine its influence on the overall pooled effect. The random-effect model was chosen to gather data, regardless of heterogeneity. Publication bias was assessed with funnel plot whenever it was possible due to the limited number of included studies in each analysis. No publication bias was detected in any analysis.

Risk of bias assessment

The ROBINS-I (risk of bias in non-randomized studies of interventions) assessment tool was used to detect bias in non-randomized included studies [21]. For the assessment of randomized clinical trials, the RoB 2 tool (risk of bias tool for randomized trials) was employed. The same reviewers performed the risk of bias assessment, also independently. Bias analysis was performed using specific criteria, i.e. patient selection (randomization), treatment blinding (if the patient was blinded to the treatment received), evaluation blinding (if the examiner was blinded to the treatment received by the patient) and statistical analysis. The risk of bias for each study was analyzed and classified according to the four pre-established domains. The overall risk of bias for the result is determined by the least favorable assessment among all the bias domains [22].

Results

A total of 353 references were found, 33 of which were duplicates. After analysis of titles and abstracts, 236 references were excluded for not being within the scope of this review. Then, 84 full texts were analyzed, and 72 additional references were discarded. Finally, 12 studies were selected and included in this review, totalizing 712 patients (Fig. 1).

Fig. 1
figure 1

Flow chart of the studies selection and inclusion in the systematic review according to the PRISMA guidelines.

Full size image

Nine studies were randomized clinical trials [23,24,25,26,27,28,29,30,31] and three non-randomized clinical trials [32,33,34]. Only 5 studies were blinded [23, 25, 27, 32, 33].

The selected clinical investigations focused on individuals presenting oral ulcers, which were either of traumatic origin (caused by prosthetic trauma, mechanical ventilation, and palatal wounds resulting from gingival graft donor sites) (n = 8 studies); or were associated with autoimmune diseases such as OLP and recurrent aphthous ulcers (n = 4). Among the studies analyzed, researchers applied topical ozone therapy to oral ulcers using three different methods: ozone gas (n = 7), ozonated oil (n = 4), and ozonated water (n = 1). Of the clinical trials included, 10 studies compared ozone therapy with a placebo group while 5 studies compared ozone therapy with the application of topical corticosteroids and/or laser photobiomodulation.

The sample size in the studies assessing the application of O3 in traumatic ulcers ranged from 7 [31] to 150 participants [33] (Table 1). The frequency, concentration, and method of application of ozone therapy varied among the studies based on their respective methodologies. Seven studies included both a test group and a placebo group (Table 1). Isler et al. [27] and Uslu [30], apart from the placebo group, included an active control group, in which laser photobiomodulation was applied to the surgical wound. The study by Tualzik et al. [31] compared the ozone group only with laser photobiomodulation.

The studies that investigated autoimmune lesions employed different protocols for the application of ozone therapy, and the sample size ranged from 51 [28] to 138 participants [24] (Table 1). Three studies compared the group that received ozone with a placebo group [24, 28, 32], whereas two other studies also included active treatment groups for comparison [24, 26].

Analysis of symptom resolution and analgesic consumption

The pain levels caused by ulcers were evaluated using the VAS scale, as well as the pain scale before, during, and after different treatments. The follow-up time for pain analysis varied from 13 days [29] to 3 months [24], and a total of 634 patients were included in the analyses.

In summary, the meta-analysis demonstrated that ozone therapy played a significant role in reducing pain in patients when compared to placebo (RR=-3.91; 95% CI -5.54 to -2.29; I2=97%). Subgroup analysis regarding the etiology of the ulcers showed that ozone therapy was significantly superior to placebo when the origin of the ulcer was traumatic (RR=-6.30; 95% CI -9.33 to -3.28; I2=96%). In autoimmune ulcers, ozone treatment was not statistically superior (RR=-2.51; 95% CI -5.16 to 0.14; I2=98%) (Fig. 2A). High heterogeneity was observed in both subgroups. In the subgroup analysis of traumatic ulcers, exclusion of Isler et al. [27] and Taşdemir et al. [29] studies reduced the I2 from 96 to 0%, which also reduced the effect of ozone therapy (RR = 1.29, 95% CI -1.6 to -0.95), but it was still statistically significant (p < 0.00001). Excluding individual studies did not result in improved heterogeneity in the analysis of the autoimmune ulcers subgroup.

Fig. 2
figure 2

Meta-analysis of ozone therapy efficacy in reducing oral ulcers symptoms. (A) Meta-analysis of visual analogue scale (VAS) scale analysis comparing the ozone therapy group versus placebo, and subgroup analysis regarding ulcer etiology (traumatic and autoimmune). (B) Meta-analysis of visual analogue scale (VAS) scale analysis comparing the ozone therapy group versus positive control, and subgroup analysis regarding active treatment (traumatic and autoimmune). (C) Meta-analysis comparing the number of analgesics consumed by patients of the ozone therapy group versus placebo

Full size image

When comparing ozone therapy specifically with topical corticosteroid [24, 26], no statistically significant difference was detected between the interventions (RR = 0.41; 95% CI -0.38, 1.20; I2=75%). Regarding the comparison of ozone therapy with laser photobiomodulation [24, 27, 30, 31], the analysis showed that the treatments were also equivalent and did not present a statistically significant difference (RR=-0.31; 95% CI -1.38, 0.77; I2=87%) (Fig. 2B). Excluding the study by Isler et al. [27] reduced the heterogeneity of this subgroup to I2=43%, and the difference between the treatments remained not significant (RR = 0.26, 95% CI -0.27, 0.78, p = 0.34).

Two studies [27, 29] recorded the number of analgesics (100 mg Flurbiprofen) consumed by patients after gingival graft surgery. Traumatic ulcers resulting from palatal wounds in gingival graft donor sites were treated with ozone gas and compared to a placebo in a total of 54 patients. The meta-analysis results indicate no statistically significant difference in analgesic consumption between the placebo and ozone therapy groups (RR=-1.69; 95% CI -4.30, 0.93; I2=93%) (Fig. 2C).

Wound size analysis

Wound healing was analyzed using different methodologies to measure the ulcerated area. Researchers used digital cameras, digital calipers, and periodontal probes. The follow-up time for ulcers varied from 14 days [27] to 28 days [23], and a total of 393 patients were evaluated.

The five studies that evaluated the size of the lesions after the application of ozone therapy compared to a placebo group (Fig. 3A) showed that O3 improves reduction of size of the ulcer, accelerating the healing process (RR=-0.96; 95% CI -1.54,-0.39; I2=82%). The high heterogeneity was caused by the studies of Patel et al. [23] and Al-Omiri et al. [32]. However, excluding these studies from the analysis did not cause significant changes in the RR, and ozone therapy still showed better performance in reducing the size of the ulcer (RR=-0.44; 95% CI -0.71, -0.17; I2=0%). The subgroup analysis (Fig. 3A) regarding the different etiologies of ulcers showed no statistical difference between them (p = 0.61).

Fig. 3
figure 3

Meta-analysis of ozone therapy efficacy in wound healing of oral ulcers. (A) Meta-analysis of reduction of wound size comparing the ozone therapy group versus placebo, and subgroup analysis regarding ulcer etiology (traumatic and autoimmune). (B) Meta-analysis of ulcera duration (days) comparing the ozone therapy group versus placebo. (C) Meta-analysis of Thongprasom scale to assess oral lichen planus (OLP) lesions, comparing ozone therapy and placebo and active treatments

Full size image

Ulcer duration

The duration of ulcers was evaluated by three studies [32,33,34], involving a total of 328 patients. The duration of the ulcer was measured in days by recording the time taken for complete healing and disappearance of the lesions. Two studies employed ozone therapy in gas form [32, 33], while one study utilized ozonated oil [34]. The meta-analysis results demonstrated that ozone therapy led to a shorter healing time and disappearance of lesions when compared to the placebo (RR=-1.53; 95% CI -1.78, -1.29; I2=0%), as depicted in Fig. 3B.

Thongprasom scoring system

Three studies that assessed the effectiveness of ozone therapy in OLP ulcers used the Thongprasom scale to evaluate the clinical manifestation of the disease [24, 26, 28]. A total of 237 patients were included in the analysis. The meta-analysis showed no statistically significant difference when comparing the ozone therapy group with the placebo group (RR=-7.89; 95% CI -22.27, 6.49; I2=99%), neither when compared to positive control (corticosteroid) (RR = 0.62; 95% CI -2.03, 3.27; I2=97%) (Fig. 3C). Both analyses presented high heterogeneity, which could not be reverted by sequential exclusion of the studies.

Bias risk assessment

The non-randomized studies [32,33,34] were classified according to the ROBINS-I tool and exhibited low risk of bias in all assessed domains (Fig. 4). Randomized clinical trials were evaluated using the ROB 2 tool (Fig. 5). Three studies had uncertain bias [24, 27, 30], mainly in the domain related to deviations from intended interventions. Six studies had a high risk of bias; three studies were assessed as having a high risk of bias in the domain related to deviations from intended interventions [23, 25, 28]; two studies had a high risk of bias in the outcome measurement domain [26, 31]; one study had a high risk of bias in both the deviations from intended interventions and outcome measurement [29]. In the overall analysis, six studies [23,24,25, 27, 28, 30] presented uncertain risk of bias, while three studies [26, 29, 31] showed high risk of bias.

Fig. 4
figure 4

Methodological quality of the non-randomized studies included in the systematic review and meta-analysis using ROBINS-I assessment tool. (A) Risk of bias graph. (B) Risk of bias summary. D1, bias due to confounding. D2, bias in selection of participants into the study. D3, bias in classification of interventions. D4, bias due to deviations from intended interventions. D5, bias due to missing data. D6, bias in measurement of outcomes. D7, bias in selection of the reported result

Full size image
Fig. 5
figure 5

Methodological quality of the randomized studies included in the systematic review and meta-analysis using RoB 2 assessment tool. (A) Risk of bias graph. (B) Risk of bias summary. D1, bias arising from the randomization process. D2.1, bias due to deviations from the intended interventions (effect of assignment to intervention). D2.2, bias due to deviations from the intended interventions (effect of starting and adhering to intervention). D3, bias due to missing outcome data. D4, bias in measurement of the outcome. D5, bias in selection of the reported result

Full size image

Discussion

Oral ulcers, even when self-limiting, significantly impact patients’ quality of life by causing pain and impairing basic functions, so it is crucial to adopt effective, practical, and safe therapeutic approaches. This systematic review revealed that ozone therapy promoted the healing of oral ulcers compared to placebo. Similar outcomes were observed when comparing ozone therapy to corticosteroid therapy and laser photobiomodulation.

Ozone therapy reduced the size of oral ulcers and accelerated the healing process compared to a placebo. This finding is consistent with a study by Zhang et al. [35], which showed that topical ozone therapy increased collagen deposition in ulcerated tissues of diabetic patients. Additionally, several studies have indicated that ozone therapy may be associated with increased levels of growth factors such as VEGF, TGF-β, and PDGF, which are crucial for the healing of gingival wounds [35, 36]. Consistent with our findings, a systematic review by Wen et al. [37] on the use of ozone therapy for the treatment of chronically refractory wounds and ulcers related to diabetic foot, reported that ozone therapy accelerated the healing process compared to placebo. However, the application of ozone therapy did not result in a significant difference compared to the use of topical and systemic antibiotics in the complete healing of diabetic ulcers.

Regarding pain reduction, previous studies have demonstrated that ozone therapy can decrease postoperative symptoms after third molar extraction [38, 39]. In this analysis, it was observed that ozone therapy resulted in more significant pain reduction compared to placebo. These findings are in accordance with the review by Sen [17], which showed that ozone may increase tissue microcirculation, leading to a reduction in the concentration of nociceptive factors, as well as decrease the activation of inflammatory pain mediators.

An animal model study showed that both laser photobiomodulation and ozone therapy effectively improved inflammation and facilitated healing of oral surgical wounds [40]. Our study confirmed the pain-reducing effects of laser photobiomodulation and ozone therapy in patients, despite conflicting results in other studies [41, 42]. However, ozone therapy and topical treatments offer greater versatility than laser application. Photobiomodulation therapy requires periodic consultations, increasing costs, whereas self-administration of ozonated oil or corticosteroids by patients promotes treatment adherence and cost reduction [43,44,45].

Ozone therapy protocols, similar to laser photobiomodulation studies for oral ulcer healing, exhibit variable protocol presentation, methods, and application frequency. Wen et al. [37] found no significant variation among different types of ozone applications. Huth et al. [46] and Kazancioglu et al. [38] reported that ozonated water showed lower cytotoxicity compared to the gaseous form, and maintained optimal biological characteristics for oral application. However, authors recommend ozonated oil to enhance the stability, handling, storage, and treatment efficacy and safety, by increasing the half-life of oxygenated agents [47, 48]. The limited number of studies in this review precluded subgroup analysis of ozone therapy application forms, preventing the determination of the optimal therapeutic strategy based on the study’s findings.

Ozone therapy was found to be equally effective as laser photobiomodulation and topical corticosteroids for traumatic and autoimmune ulcers. In the study conducted by Roopa and Jha (2022), ozone therapy combined with topical corticosteroid showed effectiveness in reducing pain and Thongprasom scores, as well as in decreasing wound size, when compared to the combination of topical corticosteroid and placebo treatment in erosive OLP [49]. It is important to note that corticosteroid therapy, even in topical formulation, is frequently associated with significant adverse effects, such as oral candidiasis, dysgeusia, burning sensation, mucosal tissue atrophy, and gastrointestinal disturbances [43, 50]. In contrast, no adverse effects were reported by patients undergoing ozone therapy in the included studies.

Notwithstanding, based on information derived from the literature, ozone gas is acknowledged as both toxic and mutagenic [155152]. The primary adverse effects manifest upon interaction with the respiratory and mucosal pathways, and the severity of these effects is contingent upon the concentration of the gas and the duration of exposure [52]. Moreover, in the comprehensive review conducted by Serra et al. [53], the prominent adverse effects observed in various studies involving ozone therapy encompassed abdominal distension, hypoesthesia, paresthesia in the lower limbs, transient exacerbation of pain, mild exacerbation reactions, and a sensation of fainting.

As elucidated by Sagai & Bocci [15], ozone therapy is contraindicated for individuals exhibiting a substantial deficiency of glucose-6-phosphate dehydrogenase (G-6PD), pregnant women, patients undergoing ACE inhibitor treatment, those with hyperthyroidism, thrombocytopenia, and severe cardiovascular instability. Hence, the recommendation is that ozone therapy be appropriately indicated and administered by a qualified professional capable of managing any emergent situations, given that delayed intervention may lead to fatal consequences [1552].

Among the limitations of the present analysis, the lack of available scientific evidence and the quality of the included studies stand out. There was a limited number of double-blind randomized clinical trials, and 6 of 9 studies presented high risk of bias. For future studies, it is suggested that double-blind randomized clinical trials be conducted with absent or properly controlled confounding variables. Additionally, the heterogeneity of ozone protocols and the different forms of application used with the different modalities of ozone therapy, as well as the presentation of the results, should be standardized to facilitate comparison.

In conclusion, this systematic review and meta-analysis showed a limited and high-risk of bias evidence that ozone therapy may be a viable option for reducing the size of oral ulcers and accelerating healing compared to placebo. Considering the available evidence, ozone therapy is as effective as laser photobiomodulation and topical corticosteroid therapy for both traumatic and autoimmune ulcers. Ozone therapy also demonstrated greater efficacy in reducing painful symptoms compared to placebo, and similar results when compared to active treatments. The evidence is still insufficient to assert the most effective form of ozone application. Thus, the dentist should carefully analyze the case and choose the most appropriate therapy, considering patient safety, treatment feasibility, and improvement in quality of life.