Matrix metalloproteinases-3 gene-promoter polymorphism as a risk factor in oral submucous fibrosis: A systematic review and meta-analysis

Abstract

Oral submucous fibrosis (OSF) is a progressive potentially malignant disorder. Matrix metalloproteinase-3 (MMP3) regulates extracellular matrix remodelling, and promoter polymorphisms may influence OSF susceptibility. This systematic review and meta-analysis evaluated the association between MMP3 expression, promoter polymorphism, and OSF risk. Electronic databases (PubMed, Cochrane, Google Scholar, ScienceDirect) were searched for studies published between 2000 and 2023, assessing MMP3 expression or promoter polymorphism in OSF. Four eligible studies were analysed using RevMan 5.4. Pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. Heterogeneity was assessed using I² statistics, and publication bias using funnel plot asymmetry and Egger’s test. No significant association was observed between OSF risk and the homozygous 5A/5A genotype (OR = 2.34, 95% CI 0.63–8.71, P = 0.21, I² = 0%) or homozygous 6A/6A genotype (OR = 1.23, 95% CI 0.20–7.56, P = 0.83, I² = 89%). The heterozygous 5A/6A genotype showed a significant association with increased OSF susceptibility (OR = 2.75, 95% CI 1.63–4.64, P = 0.0002, I² = 0%). Egger’s test indicated no publication bias. All studies demonstrated acceptable methodological quality. The MMP3 -1171 5A/6A promoter polymorphism is significantly associated with nearly three-fold increased OSF risk, whereas homozygous genotypes show no significant association. These findings suggest a role of MMP3 promoter variation in OSF pathogenesis. Larger, well-designed studies are required to validate these results and assess their diagnostic and preventive value.

INTRODUCTION

Oral submucous fibrosis (OSF) is a chronic, progressive, potentially malignant disorder characterized by excessive extracellular matrix (ECM) deposition and fibrosis of the oral mucosa.[1] Areca nut chewing remains the most established etiological factor, but genetic susceptibility influences disease onset and progression.[2,3] Among the molecular pathways involved, matrix metalloproteinases (MMPs) regulate ECM turnover, and their dysregulation contributes to OSF pathogenesis.[4,5,6,7]

MMP3 (stromelysin-1) plays a key role in degrading multiple ECM components and modulating inflammatory mediators.[8] A functional single-nucleotide polymorphism (SNP) in the MMP3 promoter region (-1171 5A/6A, rs35068180) alters transcriptional activity, wherein the 5A allele confers higher expression than the 6A allele.[9] This polymorphism has been associated with several fibrotic and malignant conditions, suggesting its potential relevance to OSF development.[10,11,12,13]

Despite increasing evidence of MMP3 involvement in OSF, findings on the association between the 5A/6A promoter polymorphism and OSF susceptibility are inconsistent.[14,15,16,17,18,19,20,21]

No prior systematic review or meta-analysis has synthesized available data. Therefore, this study aims to evaluate the association between the MMP3 -1171 5A/6A polymorphism and OSF risk, providing quantitative evidence to clarify its potential role as a genetic susceptibility marker.

METHODOLOGY

Protocol and Registration: The present systematic review was registered with the International Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42024541625. The research protocol was designed according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines 2020 [Supplementary File 1 ^{(209.6KB, pdf)} ]. No deviations occurred from the registered protocol with respect to the research question, eligibility criteria, data extraction strategy, or statistical methodology.

Focused Question: Among individuals exposed to the MMP3 -1171 5A/6A promoter polymorphism, does the presence of specific genotypes increase the risk of developing oral submucous fibrosis compared with those without the polymorphism?

Eligibility criteria

Inclusion criteria

1. Study design: Human observational studies, including retrospective or prospective case–control studies and cohort studies. Cross-sectional genetic association studies were also eligible if genotype frequencies were reported.

2. Population: Participants with clinically and/or histopathologically confirmed oral submucous fibrosis (OSF) based on accepted diagnostic criteria.

3. Exposure: Assessment of MMP3 promoter polymorphism (-1171 5A/6A, rs35068180) with validated genotyping methods (e.g., PCR-RFLP, ARMS-PCR, sequencing).

4. Comparator: Healthy individuals or non-OSF controls from the same geographic or ethnic background.

5. Outcome: Extractable genotype data (5A/5A, 5A/6A, 6A/6A) enabling odds ratio calculation.

6. Publication characteristics: Full-text original research articles published in English.

Exclusion criteria

1. Used non-human, in vitro, or animal models.

2. Were reviews, editorials, case reports, conference abstracts without genotype data.

3. Lacked confirmed OSF diagnosis or did not report clear case definition.

4. Did not specify or validate genotyping methodology.

5. Combined genotypes without providing separate frequency data.

6. Used overlapping cohorts—when duplicate datasets were identified, the study with the most complete data and largest sample size was retained.

7. Were not retrievable as full text, even after contacting authors.

Literature Search and Screening: A systematic search was conducted according to PRISMA guidelines in PubMed, Scopus, Web of Science, Cochrane Library, Google Scholar, and ScienceDirect. The last search was performed on 15 December 2024. The following query was used: (‘oral submucous fibrosis’ OR ‘OSF’) AND (‘MMP3’ OR ‘matrix metalloproteinase-3’) AND (‘polymorphism’ OR ‘5A/6A’). Filters included human studies, English language, and full-text availability. Google Scholar screening was limited to the first 300 results (2000–2024). Grey literature was explored using OpenGrey, ProQuest Dissertations, and manual screening of conference abstracts and reference lists [Table 1]. Records were deduplicated in Mendeley and screened independently by two reviewers, with disagreements resolved by consensus.

Table 1.

Electronic databases and search strategies

Database	Search strategy	Findings
PubMed	#1 (((((Matrix metalloproteinases 3) OR (MMP3) OR (mmp3) OR (Stromelysin-1) OR (Stromelysin1)	31,658
	#2 ((Oral submucous fibrosis [MeSH Terms]) OR (OSF) OR (OSMF)	5,873
	#3—(#1) AND (#2)	17
Cochrane	#1 (matrix metalloproteinases 3):ti, ab, kw OR (mmp3):ti, ab, kw OR (MMP3):ti, ab, kw (Word variations have been searched)	32
	#2 (Oral submucous fibrosis):ti, ab, kw OR (OSF):ti, ab, kw OR (OSMF):ti, ab, kw (Word variations have been searched)	241
	#3—#1 AND #2	0
Google scholar	Matrix metalloproteinases 3, Stromelysin-1, Oral submucous fibrosis, OSF, OSMF	13
Science Direct	Matrix metalloproteinases 3, Stromelysin-1, Oral submucous fibrosis	15

Study Selection: A total of 45 records were identified through database and grey literature searching. After removing eight duplicates, 37 unique studies underwent title and abstract screening by two independent reviewers. Thirty studies were excluded at this stage for the following reasons: unrelated disease/condition (n = 9), no assessment of MMP3 or MMP promoter polymorphisms (n = 7), in vitro or animal studies (n = 5), review or commentary articles (n = 4), non-English or unavailable full text despite author contact (n = 3), and duplicate conference abstracts (n = 2). Seven articles were retrieved for full-text assessment, after which three were excluded: one due to unavailability of the full text despite attempts to contact the authors, one because the study design or reported outcomes did not meet eligibility criteria (e.g., no extractable genotype frequency data), and one due to insufficient or irreconcilable baseline data required for meta-analysis. Ultimately, four studies were included in the qualitative and quantitative synthesis [Table 2]. All selection decisions and exclusion reasons were documented in the screening log, and disagreements were resolved by consensus or third-reviewer adjudication.

Table 2.

Prisma flow diagram

Data Collection: Data extraction was performed using a pre-piloted standardized form developed according to PRISMA guidelines. Two independent reviewers extracted data from each eligible study, including author name, year of publication, study design, geographic location, sample size, diagnostic criteria for OSF, genotyping method, allele/genotype frequencies, and reported effect estimates. Extracted values were cross-checked for accuracy, and discrepancies were resolved through discussion; unresolved conflicts were adjudicated by a third reviewer. When required data were missing or unclear, study authors were contacted via email for clarification. Only studies with extractable genotype counts were included in the final synthesis [Supplementary File 2]. All extracted data were recorded in duplicate to minimize transcription bias.

Supplementary File 2: Raw data extraction sheet

0	1st Author	Year	Aim/Objective	MMPs	Study Group	Control Group	Age	Gender	Country Of Study Setting
1	H.-F. Tu	2005	to clarify whether the functional nucleotide polymorphism with different promoter activity of MMP3 related to the susceptibility and the disease progression of OSCC and OSF.	1171 5A-6A polymorphisms in the promoter (MMP3)	OSCC (n = 150), OSF (n = 71)	Healthy- 98	OSCC-51.2±9.7, OSF-38.8±10.6, Control-48.0±9.0	NM	India
2	Ajay K Chaudhary gs	2010	to find out the association of the -1171 5A->6A polymorphism in the MMP-3 gene in patients suffering from OSMF, HNSCC and healthy controls.	1171 5A->6A polymorphism in the MMP-3	101 cases of OSMF, 135 of HNSCC	Healthy- 126	43.4±14.8 years	OSF-78m/23f, HNSCC-114m/21f, Control- 105m/21f.	India
3	Prajakta R gs	2016	MMP3 is crucial in connective tissue remodeling. Promoter regions are specific segments of DNA that control the rate of mRNA synthesis. Thus, polymorphism in the (5A/6A) promoter region of the MMP3 gene results in different transcriptional activities and is related to susceptibility in diseases. To detect polymorphism in the MMP3 gene-promoter region in (OSF), (OSCC), and normal individuals with tobacco and areca nut habit and without lesions, and to correlate the genotype as a risk factor in these diseases.	MMP3	OSF-5, OSCC- 5	Normal individuals with tobacco and areca nut habit (n=5) and without (n=5)	OSF- >40-3, <40- 2, OSCC- >40- 5, Control->40- 3, <40- 7	OSF and OSCC- 3/2 each, Control- 6/4 (m/f)	India
4	Preethi A. Poonja	2017	To assess the association of single-nucleotide polymorphisms, Adenosine (Insertion/ Deletion) in -1171 5A > 6A in the MMP-3 promoter regions of patients with oral submucous fibrosis and in healthy individuals (controls).	MMP3	OSF- 30,	Control- 20	19–67 year	OSF- 28M/2F, Healthy-18M/2F	India
0	Study Design		Clinical Staging of OSMF	Histopathological Grading of OSMF	Technique	Data Analysis	Limitations or future scope	Results	Conclusion
1	Observational		NS	NS	PCR- Blood	Variants- analyzed with Fisher’s exact test. The associations between the MMP3 genotype and the risk of disease genesis were estimated by odds ratio (OR), and 95% confidence interval (CI) were calculated by unconditional logistic regression models using SPSS 12.0 (Chicago, IL, USA).	NS	The 5A genotype in MMP3 promoter was observed more frequently in OSF group than in control (P = 0.01). No significant difference was noted between OSCC and control groups on the 5A genotype frequency (P = 0.18). No association was found between 5A genotype in MMP3 promoter and site or lymph node metastasis and stage of OSCC.	The results indicated that the 5A genotype of MMP3 promoter was associated with the risk of OSF but not OSCC.
1	Observational		NS	NS	PCR- Blood	Variants- analyzed with Fisher’s exact test. The associations between the MMP3 genotype and the risk of disease genesis were estimated by odds ratio (OR), and 95% confidence interval (CI) were calculated by unconditional logistic regression models using SPSS 12.0 (Chicago, IL, USA).	NS	The 5A genotype in MMP3 promoter was observed more frequently in OSF group than in control (P = 0.01). No significant difference was noted between OSCC and control groups on the 5A genotype frequency (P = 0.18). No association was found between 5A genotype in MMP3 promoter and site or lymph node metastasis and stage of OSCC.	The results indicated that the 5A genotype of MMP3 promoter was associated with the risk of OSF but not OSCC.
2	case-control study		NS	grade I, grade II, grade III, and grade IV.	PCR-RFLP- Blood	The Chi-square test (c2 test)- difference in genotypic distribution of MMP-3 promoters between the OSMF, HNSCC and control groups. Fisher’s exact test- analysis, risk was analyzed by odds ratios (OR) and 95% Confidence Intervals (95% CI). Distributions of MMP-3 genotype promoter polymorphisms in OSMF, HNSCC and control groups- Hardy-Weinberg equilibrium.	Studies with a larger number of subjects are recommended to elucidate the genetic and polymorphic contribution of MMPs in these lesions, as this may have important role in designing future therapeutic strategies.	Analysis of MMP 3 (-1171 5A->6A) polymorphism revealed the frequency of 5A allele in OSMF, HNSCC and controls to be 0.15, 0.13 and 0.07. A significant difference was found in 5A genotype frequency between OSMF (5A genotype frequency=0.15, P=0.01, OR=2.26, 95% CI=1.22-4.20) and in controls (5A genotype frequency 0.07) as well as HNSCC (5A genotype frequency 0.13, P=0.03,95%CI=1.06-3.51) and controls (5A genotype frequency=0.07) In this study, 5A genotype had greater than two fold risk for developing OSMF (OR=2.26) and nearly the same in case of HNSCC (OR=1.94) as compared to controls. In patients with OSMF as well as HNSCC, the ROC analysis between the MMP-3 genotype and age, 6A/6A allele was found to be significant in patients both over and under 45 years of age; while the 5A/5A carrier alleles showed an association only in patients less than 45 years of age.	expression of MMP-3 genotype associated with the 5A alleles, it may have an important role in the susceptibility of the patients to develop OSMF and HNSCC.
3	Piolet		NS	grade I, two cases of grade II, one cases of grade III, and one case of grade IVA.	PCR- Blood	The variants were analyzed with Fischer’s exact test using EpiCalc and GraphPad Instat software	Extent to which polymorphisms in the MMP3 gene-promoter region might actually contribute to the pathogenesis of OSF is yet to be clarified. Nevertheless, prospective experimental studies on animal models with a larger sample size and cases involving OSF turning into malignancy need to be carried out to confirm the association of the MMP3 5A/6A polymorphism with OSF.	The 5A allele in the MMP3 gene-promoter region was observed more frequently in the OSF group than the control groups. No significant difference was noted between OSCC and the control groups on the 5A allele.	The results indicated that the 5A allele of the MMP3 gene-promoter region could be associated with OSF risk factor, but not OSCC, in an Indian population.
4	Observational		Khanna et al classification into four groups	NS	PCR- Blood	A statistical analysis was done using Chi-square test and Fisher’s exact test to assess the frequency and association of the alleles in the case-control group	Further study in the field is required with a larger sample size to highlight the role of the polymorphisms in the -1171 5A > 6A promoter of MMP3-gene with OSF and also their presence in healthy individuals	Total number of 5A alleles were 13 and the 6A alleles were 87. On computing the Fisher’s exact test one-tailed there was a significant association of the 5A alleles and 6A alleles with the cases and control group, P=0.046. The frequency of the 5A allele and the 6A allele in the OSF group was greater than that in the control group. However the frequency of 5A allele in OSF was higher when compared to controls	The result showed a statistical significance difference between the duration of habit and disease severity with polymorphisms. The result also showed a higher frequency of the 5A allele in the study group as compared to the controls.

Assessment of Risk of Bias: The risk of bias in the included studies was assessed using the Newcastle–Ottawa Scale (NOS), which employs a star scoring system (0–9 stars). Each study is evaluated on three broad perspectives: selection of study groups, comparability of groups, and ascertainment of exposure or outcome of interest. Item 5 of the NOS carries 2 stars, while the remaining items each carry 1 star. A score of 9 stars indicates the highest methodological quality, with scores ranging from 5 to 9 generally considered indicative of high quality, while scores from 0 to 4 suggest lower quality.

The results of the risk of bias assessment for the selected studies indicated that all included studies were of high quality and had a low risk of bias [Table 3].

Table 3.

Risk of bias assessment

Criteria	Item	Chaudhary 2010	Poonja 2017	Tu 2005	Zade 2016
Selection	Adequacy of case definition	*	*	*	*
	Representativeness of cases	*	*	*	*
	Selection of controls			*	*
	Definition of controls	*	*	*	*
Comparability	Comparability cases/controls	**	**	**	**
Exposure	Ascertainment of exposure	*	*	*	*
	Same method of ascertainment	*	*	*	*
	Non-response rate	NA	NA	NA	NA
Total quality score		7	7	7	7
Quality		High	High	High	High

RESULTS

Characteristics of Included Studies: Four studies were included in the systematic review. One study was a pilot case–control study,[22] while the remaining three studies were observational (case–control or analytical cross-sectional studies).[8,10,18] These studies collectively evaluated a total of 206 participants in the OSF study group, 290 participants in the OSCC/head and neck squamous cell carcinoma (HNSCC) study group, and 254 participants in the control group. The age of participants ranged from 19 to 67 years.

Detailed data from the four included studies are summarized in Tables 4 and 5.

Table 4.

Demographic details, study setting, and OSF sampling details of the included studies

Author/year of publication	Sample size Study group Control group	Age (Average-Year) Sex (M/F)	Country	Study design
Hsi-Feng Tu 2005[18]	Study group OSF—70 OSCC—150 Control group—98 (Healthy areca chewers)	Age:   OSF—38.8±10.6   OSCC—51.2±9.7   Control—48.0±9.0 Sex:   OSF—71/0   OSCC—150/0   Control—98/0	Taiwan—China	Observational
Ajay Kumar Chaudhary 2010[8]	Study group OSF—101 HNSCC—135 Control group—126	Age:   OSF—43.4±14.8   HNSCC—43.4±14.8   Control—43.4±14.8 Sex:   OSF—78/23   HNSCC—114/21   Control—105/21	India	Case–control study
Prajakta R 2016[22]	Study group OSF—5 OSCC—5 Control group—10 (Tobacco and areca nut habit—5, Without habit—5)	Age:   OSF—>40–3, <40–2   OSCC—>40–5   Control—>40–3, <40–7 Sex:   OSF—3/2   OSCC—3/2   Control—6/4	India	Piolet: Prospective case–control Study
Preethi A. Poonja 2017[10]	Study group OSF—30 Control group—20	Age: OSF—19–67   Control—19–67 Sex:   OSF—28/2   Control—18/2	India	Cross-sectional case–control study

OSF—oral submucous fibrosis; OSCC—oral squamous cell carcinoma; HNSCC—head and neck squamous cell carcinoma; NM—not mentioned

Table 5.

Technique specifications, outcome, inference, and limitations of all included studies

Author/year of publication	Technique/method of analysing	Statistical analysis	Outcomes	Inference
Hsi-Feng Tu 2005[18]	PCR— From blood—MMP3 gene-promoter (-1171 5A/6A) polymorphisms	-Variants—analysed with Fisher’s exact test. -The associations between the MMP3 genotype and the risk of disease genesis were estimated by odds ratio. -95% CI was calculated by unconditional logistic regression models using SPSS 12.0 (Chicago, IL, USA).	-The 5A genotype in MMP3 promoter was observed more frequently in OSF group than in control (P=0.01). -No significant difference was noted between OSCC and control groups on the 5A genotype frequency (P=0.18). [Table S1]	The results indicated that the 5A genotype of MMP3 promoter was associated with the risk of OSF but not OSCC.
Ajay Kumar Chaudhary 2010[8]	PCR–RFLP—From blood—MMP3 gene-promoter (-1171 5A/6A) polymorphisms	-The Chi-square test (χ2 test)—difference in genotypic distribution of MMP3 promoters between the OSMF, HNSCC and control groups. - P<0.05—statistically significant. - Fisher’s exact test—analysis. -Risk—analysed by OR and 95% CI. -Distributions of MMP3 genotype promoter polymorphisms in OSMF, HNSCC and control groups—Hardy–Weinberg equilibrium. - statistical analysis was performed using the SPSS 15.0 software package.	-Analysis of MMP3 (-1171 5A->6A) polymorphism revealed the frequency of 5A allele in: OSF—0.15 HNSCC—0.13 Controls—0.07. -A significant difference was found in 5A genotype frequency between OSF (5A genotype frequency=0.15, P=0.01, OR=2.26, 95% CI=1.22–4.20) and in controls (5A genotype frequency 0.07) as well as HNSCC (5A genotype frequency 0.13, P=0.03, 95% CI=1.06–3.51) and controls (5A genotype frequency=0.07) - 5A genotype had greater than two-fold risk for developing OSF (OR=2.26) and nearly the same in case of HNSCC (OR=1.94) as compared to controls. - In patients with OSF as well as HNSCC, the ROC analysis between the MMP3 genotype and age, 6A/6A allele was significant in patients both over and under 45 years of age; while the 5A/5A carrier alleles showed an association only in patients less than 45 years of age. [Table S2]	Expression of MMP3 genotype associated with the 5A alleles, it may have an important role in the susceptibility of the patients to develop OSF and HNSCC.
Prajakta R Zade 2016[22]	PCR—From blood— MMP3	-Variants—analysed with Fischer’s exact test using EpiCalc and GraphPad Instat software. -Differences—considered significant when P<0.05.	-The 5A allele in the MMP3 gene-promoter region was observed more frequently in the OSF group than the control groups. -No significant difference was noted between OSCC and the control groups on the 5A allele. [Table S3]	The results indicated that 5A allele of MMP3 gene-promoter region could be associated with OSF risk factor, but not OSCC, in an Indian population.
Preethi A. Poonja 2017[10]	PCR restriction analysis—From blood—MMP3 gene-promoter (-1171 5A>6A) polymorphisms	-Association of variants of case–control group and the MMP3 genotypes—analysed with Chi-Square test. -Differences considered significant when P<0.05. -Frequency of distribution and associations between MMP3 genotypes and risk of disease genesis—estimated by OR and 95% CI-calculated by Fisher’s exact test.	-Total number of 5A alleles were 13 and the 6A alleles were 87. - On computing the Fisher’s exact test one-tailed there was a significant association of the 5A alleles and 6A alleles with the cases and control group, P=0.046. - The frequency of the 5A allele and the 6A allele in the OSF group was greater than in the control group. - However, the frequency of 5A allele in OSF was higher when compared to controls. [Table S4]	The result showed a statistical significance difference between duration of habit and disease severity with polymorphisms. Result also showed a higher frequency of the 5A allele in the study group as compared to the controls.

MMPs—matrix metalloproteinases, TIMPs—tissue inhibitors of matrix metalloproteinases, SNP—single-nucleotide polymorphism, PCR—polymerase chain reaction, RFLP—restriction fragment length polymorphism, OR—odd’s ratio, CI—confidence interval, NM—not mentioned; ROC—receiver operating characteristics

Statistical Analysis: Meta-analysis was performed using RevMan 5.4 (Cochrane Collaboration). Pooled effect estimates were calculated as odds ratios (ORs) with 95% confidence intervals using the Mantel–Haenszel method. Model selection was based on clinical and methodological heterogeneity in addition to statistical heterogeneity. A fixed-effect model was applied when studies were comparable in design and population characteristics and when heterogeneity was low (I² < 50% and P > 0.10 for Cochran’s Q). A random-effects model (DerSimonian–Laird) was used when substantial heterogeneity was present (I² ≥ 50% or Q-test P < 0.10), as this model accounts for between-study variance and provides more conservative pooled estimates. Heterogeneity was quantified using I² values, interpreted as low (25–49%), moderate (50–74%), or high (≥75%) inconsistency. Publication bias was assessed using funnel plots and Egger’s regression test. All analyses were performed under a predefined protocol without post hoc model switching.

Comparison of Frequency of Genotypes between OSF and Healthy Controls:

1. Homozygous 5A Genotype (5A/5A): [Image 1—Forest Plot 5A/5A]

   • Meta-analysis included three studies with sufficient quantitative data.

   • Heterogeneity (I² = 0%) was low; hence, a fixed effects model was applied.

   • Frequency of homozygous 5A genotype did not significantly differ between OSF and healthy controls.

   • Odds ratio (OR) = 2.34; 95% confidence interval (CI): 0.63 to 8.71; P = 0.21.

   Assessment of Publication Bias:

   • Funnel plot [Supplementary Image 1 (S1) ^{(493.4KB, tif)} ] showed symmetrical distribution.

   • Egger’s test: P = 0.698, indicating no publication bias.

2. Homozygous 6A Genotype (6A/6A): [Image 2—Forest Plot 6A/6A]

   • Meta-analysis included three studies with sufficient quantitative data.

   • Heterogeneity (I² = 89%) was substantial; thus, a random effects model was applied.

   • Frequency of homozygous 6A genotype did not significantly differ between OSF and healthy controls.

   • OR = 1.23; 95% CI: 0.20 to 7.56; P = 0.83.

   Assessment of Publication Bias:

   • Funnel plot [Image S2 ^{(505.3KB, tif)} ] showed symmetrical distribution.

   • Egger’s test: P = 0.093, indicating no significant publication bias.

3. Heterozygous Genotype (5A/6A): [Image 3—Forest Plot 5A/6A]

   • Meta-analysis included three studies with sufficient quantitative data.

   • Heterogeneity (I² = 0%) was low; hence, a fixed effects model was applied.

   • Frequency of heterozygous genotype was significantly greater among OSF subjects compared to healthy controls.

   • OR = 2.75; 95% CI: 1.63 to 4.64; P = 0.0002.

   Assessment of Publication Bias:

   • Funnel plot [Image S3 ^{(493.4KB, tif)} ] showed symmetrical distribution, suggesting no potential publication bias.

   • Egger’s test: P = 0.052, confirming no publication bias.

Image 1.

Forest plot 5A/5A

Image 2.

Forest plot 6A/6A

Image 3.

Forest plot 5A/6A

DISCUSSION

This systematic review identified upregulated expression of MMP3-specific genetic polymorphism in OSF through a comprehensive literature search that yielded four relevant studies.[8,10,18,22]

OSF is a premalignant disorder induced primarily by areca nut chewing. Arecoline, a component of areca nut, increases collagen fibre production in the buccal cavity’s subepidermal layer, leading to fibrosis. MMP3 also plays a crucial role in oral cavity fibrosis. Chinu et al. reported a correlation between areca nut chewing and genetic polymorphisms associated with collagen homeostasis like MMP3 and OSF risk.[10]

This study aimed to explore the association between SNP in OSF patients and healthy controls, specifically evaluating the adenosine (insertion/deletion) in -1171 5A > 6A polymorphism in the MMP3 promoter region. The results showed associations among the 5A/5A, 6A/6A, and 5A/6A genotypes and various variables.

Chaudhary et al.[8] reported a significant association of the 5A allele with susceptibility to HNSCC, suggesting a nearly two-fold increased risk among individuals with the 5A/5A genotype compared to controls. Tu et al.[18] [Table S1] similarly found the 5A/5A genotype associated with OSF risk but not OSCC.

Table S1.

Outcomes of Hsi-Feng Tu 2005: Genotypes in MMP3 promoter related

Numbers	5A genotype frequency	5A/5A allelotype	5A/6A allelotype	6A/6A allelotype	P*	OR	95% CI
Controls—98	0.07	1 (1)	12 (12)	85 (87)		1
OSCC—150	0.10	0 (0)	21 (31)	119 (79)	0.17	1.70	0.84–3.45
OSF—70	0.15	1 (2)	19 (27)	50 (71)	0.02	2.62	1.33–7.89

P*—5A/5A and 5A/6A allelotypes vs. 6A/6A allelotype; OR—odds ratio; CI—confidence interval; OSF—oral submucous fibrosis; OSCC—oral squamous cell carcinoma

In Poonja et al.’s study, the frequency of 5A/6A or 5A/5A genotypes with 5A alleles in OSF was significantly higher compared to controls, aligning with findings from other studies. The study concluded that the 5A genotype may confer more than a two-fold risk for OSF and slightly less for HNSCC compared to controls.[8,10]

Tu et al.[18] observed a significant difference in MMP3 genotypic polymorphism between control subjects and OSF patients among male areca nut chewers, highlighting a greater than three-fold risk associated with the 5A allele.

Chaudhary et al.[8] [Table S2] used receiver operating characteristics (ROC) analysis to link age with MMP3 genotypes in OSF and HNSCC, revealing significant associations with specific alleles across different age groups. 6A/6A allele was significant in subjects both over and under 45 years of age. However, 5A/5A carrier alleles showed an association only in patients less than 45 years of age. Fisher’s exact test also revealed that patients with 5A alleles had a significant risk to develop OSF (P = 0.01, OR = 2.51, 95% CI = 1.27–4.94).

Table S2.

Outcomes of Ajay Kumar Chaudhary 2010

Groups	5A allele frequency	5A/5A Genotype	5A/6A Genotype	6A/6A Genotype	P	OR	95% CI
Controls—126	0.07	2 (1.58%)	14 (11.1%)	110 (87.3%)		1
OSF—101	0.15	3 (2.9%)	24 (23.7%)	74 (73.3%)	0.01	2.26	1.22–4.20
HNSCC—135	0.13	6 (4.44%)	23 (17.03%)	106 (78.52%)	0.03	1.94	1.06–3.51

The genotype and allele frequency of MMP3 in OSF, HNSCC and controls—[OR=odd ratio; CI=confidence interval; MMP=matrix metalloproteinase; OSF=oral submucous fibrosis; HNSCC=Head and neck squamous cell carcinoma]

In contrast, Han G et al. found no association between MMP3 SNP and OSCC progression.[23]

Zade PR et al. [Table S3] successfully amplified the MMP3 gene-promoter region in their study, confirming the presence of the 5A allele in OSF cases. They also noted variations in OSCC cases, suggesting different genetic predispositions between the two conditions. Further research is needed to elucidate these observations.[22]

Table S3.

Outcomes of Prajakta R Zade 2016: Alleles in the matrix metalloproteinase-3 promoter region

Groups	5A allele	6A allele	Two-sided P
OSF—5	4	1	0.0049 (very significant)
OSCC—5	1	4
NH—5	-	5
N—5	-	5

N—without tobacco and areca nut habit; NH—with tobacco and areca nut habit; OSCC—oral squamous cell carcinoma; - OSF—oral submucous fibrosis; Row/column association is statistically significant

With respect to the findings of the Zade PR, et al. study, it can be concluded that the 6A allele was predominantly present in OSCC, which is similar to that of normal individuals with and without a tobacco and areca nut habit. It has been reported that arecoline directly stimulates cultured mucosal fibroblasts.[24] However, individuals with a tobacco and areca nut habit still show the 6A allele, similar to that of normal individuals without the habit. It can be stated that the 6A allele is not a risk factor for OSCC.

Poonja et al.[10] found that females constituted 6.7% of the study group and 10% of the control group. The 5A/5A and 5A/6A genotypes were predominantly found in males. However, a Chi-square test indicated no significant difference in genotype distribution between [Tables S4 and S5] genders.

Table S4.

Outcomes of Preethi A. Poonja 2017: Distribution of allele frequency

Allele frequency	5A	6A	Odd’s ratio (OR)	95% CI	Fisher’s exact test (one-tailed)	Fisher’s exact test (two-tailed)	Pearson’s Chi-square test
Study G	11	49	4.27	0.84–41.37	P=0.046	P=0.07	P=0.052
Control G	2	38
Total	13	87
Frequency in OSF	84.6%	56.3%

Total number of 5A alleles—13 and the 6A alleles—87. Fisher’s exact test one-tailed—significant association of the 5A alleles and 6A alleles with the cases and control group, P=0.046. The frequency of the 5A allele and the 6A allele in the OSF group—greater than that in the control group. However, the frequency of 5A allele in OSF—higher compared to controls

Table S5.

Association of gender with 5A/5A, 5A/6A, 6A/6A in study and control groups

Groups	Gender	5A/5A		5A/6A		6A/6A
		Present	Absent	Present	Absent	Present	Absent
Study G	Male	3	25	5	23	20	8
	Female	0	2	0	2	2	0
	Total	3	27	5	25	22	8
Control G	Male	0	18	2	18	18	2
	Female	0	2	0	2	0	2
	Total	0	20	2	20	18	4
Chi-square value		0.238		0.429		0.779
P		0.626		0.513		0.377

The female population included in the study group—6.7% and in the control group—10%. The frequency of the 5A/5A, 5A/6A genotype in the study population was present only in males. On computing the Chi-square test—no significant difference present between the gender and genotypes

Chang et al.[25] discovered that arecoline not only inhibits MMP2 gelatinolytic activity but also stimulates tissue inhibitor of metalloproteinase (TIMP-1) activity in OSF. Their study suggested that MMP3 transcriptional activity might be linked to OSF development in young areca nut chewers.

Poonja et al. and other studies noted a majority of study participants aged between 19 and 34 years.[10,26,27] Individuals under 35 years were significantly more likely to develop OSF compared to older individuals. However, age groups [Table S6] did not show statistical significance concerning genotypes.[10]

Table S6.

Allele population distribution according to age

Age groups	5A/5A	5A/6A	6A/6A
19–34 years	3	4	25
35–50 years	0	2	12
51–67 years	0	1	3
P	0.40	0.793	0.811

Mean age—33.71 yr—which apparently coincides with the age distribution worldwide in terms of propensity to premalignant lesions (OSF). Association of age groups with 5A/5A, 5A/6A, 6A/6A. No significant association—between the age groups and polymorphism

In a study conducted by Seedat HA, Van Wyk CW it was interesting to note that some patients developed OSF after a short areca-exposing time along with some unexposed ones. However, other studies concluded that chewing betel nut will lead to pathological changes in the mucosa, but, neither the frequency nor the duration of the habit, are accurate predictors of the extent of these changes or when they are likely to occur.[10,21] In a study by Chaudhary et al.,[8] MMP3 gene polymorphisms in OSF and HNSCC showed significant differences between areca nut with tobacco chewers and non-chewers.

In Poonja et al.’s study,[10] 73.4% of the OSF patients consumed betel quid with tobacco, 13.3% chewed only betel quid, and 13.3% consumed tobacco with betel quid and alcohol. They analysed genotype association with these habits [Table S7] but found no significant difference. They also assessed the association of tobacco chewing duration, quantity, frequency, and OSF stages with genotypes. [Table S8] Significant associations were found with duration and OSF stage, suggesting that longer habits and disease severity may be linked to specific genotypes. However, the study acknowledged the need for a larger sample size to validate these findings and explore markers of disease progression or malignancy susceptibility further.

Table S7.

Allele population distribution based on the type of habit

Type of habit	5A/5A	5A/6A	6A/6A
Betel quid	0	0	4
Betel quid + tobacco	2	3	17
Betel quid + tobacco + smoking + alcohol	1	2	1
No habit	0	2	18
P	0.21	0.15	0.01

Association of type of habit with polymorphism. No significant difference was noted with the type of habit and 5A/5A, 5A/6A (polymorphism)

Table S8.

Association of quantity of the habit with polymorphism

Quantity (no of packets/day)	5A/5A	5A/6A	6A/6A
< 5	3	4	10
> 5	0	1	12
Controls	0	2	18
P	0.045	0.37	0.02

The association of the quantity of the product consumed was statistically significant with the 5A/5A polymorphism P=0.04

In some case–control studies, a particular genotype is not associated with the disease, while in other studies, the same genotype is associated with the disease. This unexpected result could have been due to observed linkage disequilibrium between MMP3 and the adjacent MMP1 locus. A haplotype analysis of the MMP gene cluster on chromosome 11q22, which includes the MMP10, MMP1, MMP3, and MMP13 genes, will probably provide deeper insight into the complex relationship between the MMP3 polymorphism and precancer (OSF) and cancer risk (OSCC).[6]

These studies collectively underscore the role of MMP3 promoter polymorphism in modulating MMP3 transcription levels, potentially influencing OSF and OSCC pathogenesis.

Although the 5A/6A polymorphism of MMP3 shows a significant statistical association with OSF risk, the available evidence remains insufficient to establish a causal relationship. The pooled results should be interpreted cautiously due to limited sample size, potential confounding factors, and lack of functional validation. Therefore, the findings indicate correlation, not causation, and require confirmation in larger, multiethnic studies with mechanistic analysis.

CONCLUSION

This meta-analysis identifies a significant association between the MMP3 -1171 5A/6A promoter polymorphism and oral submucous fibrosis risk, but the evidence is limited by small sample sizes, geographic restriction, and methodological variability. The association is correlational and cannot be considered causal without functional studies or adjustment for environmental factors, such as areca nut exposure. While MMP3 variation may contribute to genetic susceptibility, current data are insufficient to confirm its value as an independent biomarker. Larger, multi-ethnic studies incorporating gene–environment interaction and mechanistic validation are required.

FUTURE PERSPECTIVES

This review highlights important limitations in existing studies on the MMP3 -1171 5A/6A polymorphism and OSF. Most studies had small sample sizes, limited follow-up, and restricted participant groups. Larger, well-designed studies with long-term follow-up are needed to clarify whether this polymorphism influences OSF risk, prognosis, or malignant transformation. Future research should also include broader comparison groups, such as habitual chewers without OSF and OSF patients without risk habits, to better define gene–environment interactions. Experimental animal studies following OSF progression to malignancy would further support mechanistic evidence. Addressing these gaps could help establish MMP3 polymorphism as a diagnostic or therapeutic marker for OSF-related disease.

Conflicts of interest

There are no conflicts of interest.

Image S1

Funnel Plot 5A/5A

Image S2

Funnel Plot 6A/6A

Image S3

Funnel Plot 5A/6A

SUPPLEMENTARY FILE 1

PROSPERO PROTOCOL

Funding Statement

Nil.