Oral, gut, skin, nose and urogenital microbiota all interact with functions and factors like immunity, metabolism and nutrition (Pflughoeft & Versalovic, 2012). The number of cells in the oral bacteriome is less than 1% of the number in the colon bacteriome, but the proportional numbers of saliva and dental plaque bacteria are considerable (109 e 1011 per mL, respectively) (Sender et al., 2016). A comprehensive report of all the bacterial species found into the oral cavity is publicly available at the Expanded Human Oral Microbiome Database (EHOMD, www.homd.org) (Dewhirst et al., 2010). The analysis of about 35,000 sequences of the RNA 16 S through the Basic Local Alignment Search Tool (BLASTN, https://blast.ncbi.nlm.nih.gov/Blast.cgi) allowed to confirm the 619 species into the HOMD and identified other 434 taxa. Multi-omics approaches currently allow to investigate the polymicrobial complexity, interspecies interactions and mechanistic insights (Shokeen et al., 2021).

Bacteria are not uniformly distributed across oral surfaces but grow differently due to their metabolism, co-evolution of microbe-microbe interactions and host features (Mark Welch et al., 2019, Samaranayake and Matsubara, 2017). Several evidences showed a site-specific signature of the oral bacteriome (Shailaja et al., 2013),(Douglas et al., 1993, Mitchell, 2011),(Eren et al., 2014, Kazor et al., 2003). The complex design includes a multi-level structure of the oral biofilm with bacteria selectively present in the basal, intermediate, and surface levels interacting each other (Zijnge et al., 2010). The site-specific hypothesis (Socransky and Manganiello, 1971, Zhou et al., 2013) is supported by the heterogeneity of anatomical and by the mutual influence bacteria have on one another. The Human Microbiome Project (HMP) pointed out a main core of taxa, diverse across oral sites but maintained across individuals, belonging to several families and phyla (Lloyd-Price et al., 2017). The variable components of the bacteria integrate this main core and account for the diverse genetics and lifestyle of each individual (Zarco et al., 2012); the resulting "oral bacteriome imprinting" represents a possible personalized risk level for oral diseases.

In addition to the inherent variability, oral microbiota may vary due to sampling and analytical procedures, like stimulated and unstimulated saliva, oral rinse, oral swab, scraping, cytobrush, and biopsy (Belstrøm et al., 2016, Vogtmann et al., 2019). Saliva represents all ecological niches and can readily be analysed for biomarkers; however, unstimulated and stimulated samples differ in volume and composition. If a non-invasive, site-specific procedure is required, oral brushing should be preferred over saliva sampling (because non-specific) or biopsies (because invasive). Oral rinse consists of a slightly different bacterial composition than saliva, with a greater alpha-diversity (G. Yu et al., 2017). Salivary bacteriome is similar to that of the tongue, suggesting that most of the Operational Taxonomic Units (OTU) greatly abundant into the saliva derive from the tongue (Hall et al., 2017).

The different techniques for nucleic acids extraction also affect the subsequent bacteriome analyses (Hiergeist et al., 2016),(Abusleme et al., 2013). For what concerns sequencing, the gene 16 S rRNA is the most used: after a polymerase chain reaction (PCR) amplification, the nine hypervariable regions V1-V9 (30–100 bp) allow to determine species. Next generation sequencing (NGS) allows to sequence the whole genome, and metagenomic sequencing allows to determine all the genes into a community while informing on the functions. Taxonomy is a crucial element of microbiome research: OTU can include more species and one species can be present in more than one OTU. Among the metrics, richness refers to the number of OTU while diversity indexes (such as Shannon's and Simpson's) estimate the true diversity.

The term OPMD has been introduced in 2007 by the World Health Organization (WHO), overcoming the distinction between lesions and conditions, and it has recently been updated (Warnakulasuriya et al., 2021). The term OPMD refers to the fact that, even in patients with a definite lesion, malignant transformation may occur elsewhere in the oral cavity rather than on the lesion, and that its presence is not necessary but simply possible. Leukoplakia, erythroplakia, oral lichen planus, oral submucosal fibrosis, actinic cheilitis, palatal lesions from reverse cigar smoking, discoid lupus erythematosus, and some hereditary disorders such as congenital dyskeratosis and Fanconi anemia are all covered under the umbrella term OPMD. Among them, lichen planus and the different forms of leukoplakia have received the most attention in the field of studies on the oral microbiota.

Leukoplakia (OLK) refers to “predominantly white plaque of questionable risk, having excluded (other) known diseases or disorders that carry no increased risk for cancer” (Warnakulasuriya et al., 2007). Prevalence is about 2%, higher as age increases (Maymone et al., 2019). Leukoplakia is usually diagnosed in the fifth-sixth decade of life. Incidence has been reported as 1.1 - 2.4 / 1000 on males and 0.2 - 1.3 / 1000 per year (Bewley & Farwell, 2017). Malignant transformation is more probable in case of non-homogeneous lesions (14.5% vs 3.0%) (Warnakulasuriya & Ariyawardana, 2016). The erythematous component has been highlighted as a sign of malignant degeneration (Bánóczy & Sugár, 1975). Erythroplakia, as a "fiery red patch that cannot be characterized clinically or pathologically as any other definable disease" (Warnakulasuriya, 2018), often manifest as an erythematous lesion, with well-defined margins, velvety and often symptomatic, in any site of oral and oropharyngeal cavity. It is usually diagnosed in the sixth-seventh decade of life, with no gender imbalance. Prevalence is 0.02-to-0.83% (Villa et al., 2011). Malignant transformation rates are high ranging from 14% to 50% (Wetzel & Wollenberg, 2020). The association of leukoplakia and erythroplakia, i.e., white areas interspersed with red areas, defines erythroleukoplakia.

Proliferative verrucous leukoplakia (PVL) is a distinct subset of OLK characterized by multifocal, recurrent, and exophytic lesions, with a high rate of malignant progression. PVL is more frequent on females (4:1) older than 60 years. PVL lesions are initially manifested asymptomatically, with small and defined white plaque that slowly grow and affect multiple surfaces of oral mucosa. Such lesions are going to expand and become more warty and exophytic (Speight et al., 2018). Gingiva is the most represented site, malignant rate ranges from 63.3% to 100% with an average time of 5–6 years, recidivism after ablation treatment is highly probable (85%) (Pentenero et al., 2014).

Oral lichen planus (OLP) is a chronic, non-remissive disease that presents multiple co-morbidities (Parashar, 2011). Prevalence is 0.5-to-2%, males to females ratio 2:1, and age of onset 20–60 years (Eisen et al., 2005, Giuliani et al., 2019). OLP often manifests bilaterally on the buccal mucosa, despite other oral mucosal sites can be affected (McParland, 2016). Among the subtypes, reticular and erosive/ulcerative ones can be mentioned. OLP is worsened by stress (De Porras-Carrique et al., 2022) and periodontopathogenic bacteria (Liu et al., 2021). It is important to distinguish OLP from oral lichenoid lesions, since the latest are featured by a higher malignant transformation rate (2.5–3.2% vs 0.9–1-1%) (Aghbari et al., 2017, Fitzpatrick et al., 2014, Giuliani et al., 2019). Lesions affecting tongue, and erosive subtypes are more likely to transform (Aghbari et al., 2017, Giuliani et al., 2019).

Site, size, age of onset, sex, and lifestyle habits are all risk factors for malignant transformations of OPMDs. Despite tobacco and alcohol use are known to favor the onset of OPMDs (González-Álvarez and García-Pola, 2022, Kusiak et al., 2020), other factors are required to observe a malignant transformation.

Microbiota and oral cancer interact into a complex and dynamic microenvironment (Pignatelli et al., 2022). Polymicrobial synergy determines a microenvironment favoring tumoral degeneration through the Microbe-Associated Molecular Patterns (MAMP), Pathogen-Associated Molecular Patterns (PAMP) and Damage-Associated Molecular Patterns (DAMP), with a microbiota - cancer axis recently defined "oncobiome" (D’Antonio et al., 2022, Thomas and Jobin, 2015). Epithelial defects, genotoxicants and bacterial pro-tumoral metabolites importantly favor the onset of cancers (Pignatelli et al., 2021, Wells and Coyne, 2018).

The "keystone-pathogen hypothesis" holds that certain low-abundance oral bacteria interact through several pathways such as heterolactic fermentation by promoting growth of other bacteria thereby remodeling into a dysbiotic microbiota (Hajishengallis et al., 2012). Oral dysbiosis in turn promotes carcinogenesis and bacteria then promote the homeostasis of tumor microenvironment. Microbial communities can create a harmful cooperation with risk factors such as tobacco and alcohol use thereby favoring cancer (Fig. 1).

Does the microbiota account for cancer prognosis? The hypoxic nature of tumor microenvironment, the altered host immunity, and the purine production by the necrotic zone all promote the growth of selective bacterial species (Sonalika et al., 2012). In turn, bacterial species such as F. nucleatum promotes the advancement of oral squamous cell carcinoma (OSCC) (Perera et al., 2016, Pignatelli et al., 2023). Lactic acid bacteria have a Janus-faced nature, both promoting and protecting, by reducing pH: one the one hand, they reduces the environmental pH thereby favoring the growth and diffusion of cancer cells (La Rosa et al., 2020), on the other hand the lactic acid favors apoptosis, immunoprotection, and expression of tumor-suppressing genes (L. Zhong et al., 2014). Inconsistent results of research studying the involvement of bacteria in cancer development and progression can be attributed to heterogeneity in patient, tumor zones, and follow-up period. Heterogeneity also exists in the associations of OPMDs with the bacteriome, still limiting the understanding of possible pathophysiological pathways.

Within this background, observational reports of the associations between bacteriome of patients with OPMD and whether dysbiosis affect the course of diseases may vary across studies, due to differences in sampling methods and sites, sequencing methods, disease stage, controls and patients characteristics. The main aim was to systematically evaluate the bacteriome of patients with oral potentially malignant disorders such as leukoplakia, erythroleukoplakia, erythroplakia, oral lichen planus. The secondary aim was to depict whether oral dysbiosis affect etiopathogenesis and/or prognosis of the mentioned diseases. Adapting the model PICO (Patient/Population, Intervention/Exposure, Comparison/Control, Outcome) and its numerous variants (Methley et al., 2014), the research question has been structured as follows, and further checked with the FINER (Feasible, Interesting, Novel, Ethical, Relevant) model (Cummings et al., 1988):

P (Patient): oral potentially malignant disorders

C (Comparison): healthy controls and/or other diseases

O (Outcome): bacterial species

S (Study type): observational study