Diagnostic and Therapeutic Approaches in Periodontology: From Traditional Concepts to Modern Innovations

1. Introduction

Periodontitis is a highly prevalent chronic inflammatory disease that represents a major burden for both oral and systemic health. It is characterized by a dysbiotic subgingival microbiome and a dysregulated host immune response, leading to progressive destruction of the tooth-supporting tissues. In routine practice, diagnosis still relies mainly on clinical parameters and two-dimensional radiographic assessment. Although these approaches remain indispensable, they largely reflect cumulative tissue damage and provide limited insight into ongoing disease activity or future progression.

Recent advances in molecular biology and digital technologies have substantially reshaped the current understanding of periodontitis as a complex and multifactorial condition. The identification of biomarkers in saliva and gingival crevicular fluid, together with the development of omics-based approaches and microbiome profiling, has enabled a more detailed characterization of host–microbe interactions. At the same time, progress in digital imaging and artificial intelligence has improved the potential for more accurate detection, classification, and risk assessment.

However, the clinical translation of these innovations remains uneven. Key challenges include the lack of standardized and validated biomarkers, limited external validation of emerging technologies, and the practical difficulties of integrating molecular and digital data into routine clinical workflows. As a result, their impact on everyday periodontal practice is still evolving.

In this context, this review provides an updated overview of current diagnostic and therapeutic approaches in periodontology, with particular emphasis on the transition from conventional methods to emerging molecular and digital strategies, and their implications for precision-based and individualized patient care. This narrative review was conducted following a structured literature search across major biomedical databases, including PubMed, Scopus, and Web of Science. The search strategy combined keywords and Medical Subject Headings (MeSH) related to “periodontitis,” “diagnosis,” “therapy,” “biomarkers,” “digital imaging,” and “precision medicine”. Boolean operators (AND, OR) were applied to refine the search, and filters were set to include articles published between 2015 and 2025 in English. Special emphasis was placed on studies addressing emerging diagnostic technologies (biomarkers, omics, microbiome profiling, digital imaging, artificial intelligence) and novel therapeutic approaches (host-modulation therapies, regenerative strategies, precision-based interventions). The methodological quality of included studies was appraised narratively, considering study design, sample size, diagnostic criteria, and potential confounders.

2.1. Conventional Clinical Parameters for Periodontal Diagnosis

The periodontal clinical examination constitutes the cornerstone of traditional diagnosis and remains the reference standard against which emerging diagnostic modalities are compared. Through a set of systematic linear measurements and detailed soft tissue assessments, clinicians obtain a comprehensive representation of the inflammatory status, the extent of periodontal support loss, and the integrity of the tooth-supporting apparatus [1,2,3]

From an operational perspective, clinical parameters enable the identification of reversible and irreversible inflammatory conditions, the quantification of periodontal support loss in terms of severity and extent, the monitoring of post-therapeutic changes during active treatment and maintenance phases, and the assignment of disease stage and grade according to the framework proposed by the 2017 World Workshop and implemented in the 2018 Periodontal Classification System [4,5].

2.1.1. Probing Depth (PD)

Probing depth (PD), measured from the free gingival margin to the base of the gingival sulcus or periodontal pocket, reflects the current inflammatory status and the position of the junctional epithelium. Under periodontal health conditions, PD typically ranges between 1 and 3 mm; greater values indicate apical migration of the junctional epithelium and loss of attachment, making it essential to distinguish between true periodontal pockets and pseudopockets [2].

The force applied during periodontal probing is a critical determinant of diagnostic reproducibility. Excessive pressure may penetrate the ulcerated epithelium and result in overestimation of probing depth, whereas insufficient force may lead to underestimation. A probing force of approximately 0.25 N is therefore recommended to ensure standardized and reproducible measurements [3].

Within the context of the 2018 Periodontal Classification, probing depth contributes to defining case complexity in the staging framework. Probing depths ≥6 mm, particularly when associated with furcation involvement or increased tooth mobility, are indicative of advanced stages of periodontitis [6]

2.1.2. Clinical Attachment Level (CAL)

Clinical attachment level is the parameter that most accurately reflects cumulative periodontal tissue destruction and is considered one of the most reliable indicators of cumulative periodontal tissue destruction in longitudinal studies (Heitz-Mayfield, 2024). It is defined as the distance from the cemento–enamel junction (CEJ) to the base of the gingival sulcus or periodontal pocket [2]. CAL integrates probing depth and the position of the gingival margin relative to the CEJ and is calculated using the formula: CAL = PD ± (GM–CEJ) [3].

The 2017 World Workshop on the Classification of Periodontal and Peri-Implant Diseases and Conditions established diagnostic thresholds based on CAL: Stage I (1–2 mm), Stage II (3–4 mm), and Stages III–IV (≥5 mm or tooth loss attributable to periodontitis) [4]. The extent of attachment loss and the proportion of radiographic bone loss relative to root length further refine clinical and radiographic interpretation within the staging framework [6].

2.1.3. Bleeding on Probing (BOP)

Bleeding on probing (BOP) is one of the earliest and most sensitive clinical signs of gingival inflammation. A site is considered BOP-positive when bleeding occurs within 10 seconds after probe insertion using a standardized force of approximately 0.25 N. The presence of BOP reflects ulceration of the junctional epithelium and increased vascular fragility associated with inflammatory changes [7].

Although the positive predictive value of BOP for future disease progression is relatively low (approximately 30%), its negative predictive value exceeds 95%, indicating that the absence of BOP is strongly associated with periodontal stability [3]. According to the EFP S3 clinical practice guidelines, a BOP frequency below 10% of sites defines gingival health, whereas values between 10–30% and above 30% are indicative of localized and generalized gingivitis, respectively [6].

2.1.4. Furcation Involvement

Furcation involvement represents a critical clinical manifestation in multirooted teeth and reflects advanced periodontal breakdown. It is detected by horizontal probing using a Nabers probe. It is commonly classified according to the system proposed by Hamp et al. as Grade I (≤3 mm horizontal loss), Grade II (>3 mm without through-and-through involvement), and Grade III (through-and-through furcation involvement) [8].

The presence of furcation involvement significantly increases therapeutic complexity. It is considered a key diagnostic criterion for Stages III and IV in the 2018 periodontal classification, particularly when associated with deep pockets, tooth mobility, or occlusal instability [9].

2.1.5. Tooth Mobility

Tooth mobility reflects functional loss of periodontal ligament fibers and alveolar bone support. It is assessed by applying alternating horizontal and vertical forces to the tooth. It is traditionally classified according to Miller (1950) as Grade 0 (<0.2 mm), Grade 1 (0.2–1 mm horizontal movement), Grade 2 (>1 mm horizontal movement), and Grade 3 (vertical and/or rotational mobility) [3].

Mobility equal to or greater than Grade 2 indicates advanced periodontal support loss and is frequently associated with occlusal trauma or pathological tooth migration. Within the current classification framework, tooth mobility constitutes an important complexity factor for staging, particularly in Stage IV periodontitis [6].

2.1.6. Tooth Loss Attributable to Periodontitis

Tooth loss due to periodontitis represents the most severe and irreversible outcome of periodontal disease, reflecting cumulative destruction of the supporting tissues over time [1]. In contrast to probing-based parameters, which may vary with inflammatory status or treatment, tooth loss constitutes a definitive marker of disease history and long-term impact.

Within the 2018 periodontal classification, tooth loss attributable to periodontitis is a key determinant of disease staging. Stage III is defined by the loss of up to four teeth due to periodontal causes. In contrast, Stage IV involves the loss of five or more teeth, frequently associated with functional impairment or occlusal instability [10].

Accurate attribution of tooth loss is essential, as extractions related to caries, trauma, or endodontic failure should not be considered in periodontal staging. Proper classification requires integration of clinical findings, radiographic evidence, and patient history [2]. Beyond staging, periodontal tooth loss has important functional and rehabilitative implications, reinforcing the central role of prevention and long-term maintenance in periodontal care [1,10].

2.1.7. Prognostic Value

Conventional clinical parameters remain the global standard for the diagnosis of periodontitis and the assessment of therapeutic outcomes. However, their predominantly retrospective nature limits their ability to identify active disease in real time. In contemporary periodontal practice, there is a growing emphasis on integrating clinical examination with molecular biomarkers, digital radiographic analysis, and artificial intelligence–based predictive models to complement traditional assessments with information related to disease activity and future risk [6].

2.2. Two-Dimensional Radiography and Its Limitations

2.2.1. General Considerations

Modern periodontal diagnosis relies on integrating clinical parameters and radiographic information to quantify the extent and pattern of alveolar bone loss. Historically, two-dimensional (2D) radiography—primarily periapical, bitewing, and panoramic imaging—has represented a fundamental pillar for assessing periodontal support loss and documenting treatment outcomes [11]. While clinical examination identifies current inflammatory status, radiographic imaging provides an objective representation of the mineralized component of the periodontium, reflecting mainly the cumulative history of the disease rather than its immediate biological activity [2,3].

Within the framework of the 2017 World Workshop classification, radiographic bone loss constitutes an essential criterion for determining the severity of periodontitis, and its pattern and distribution complement the assignment of clinical stage and disease grade [4]. Correlation between clinical and radiographic findings supports a comprehensive diagnosis grounded in both morphological and functional evidence.

However, the projection of three-dimensional structures onto a two-dimensional plane inevitably results in loss of spatial information, leading to underestimation of bone defects, particularly on buccal and lingual surfaces. Moreover, radiographic changes typically become apparent only after 30–50% of mineral bone density has been lost [11,12,13]

Despite these limitations, 2D radiography remains the imaging modality of choice in many clinical settings due to its accessibility, low cost, reproducibility, and favorable risk–benefit ratio [14]. Its interpretation should always be contextualized within clinical findings, systemic risk factors, and evidence-based guidelines, consolidating its role as a complementary tool within the paradigm of precision periodontology.

2.2.2. Panoramic Radiography

Panoramic radiography is frequently used as an initial imaging examination due to its ability to provide a comprehensive overview of both jaws, the dentition, and adjacent anatomical structures, enabling the identification of relevant concomitant findings within a single exposure [15]. In periodontology, it is particularly useful for obtaining a general appreciation of the distribution of interproximal bone loss and for preliminary assessment during the first clinical visit or within interdisciplinary treatment approaches [11].

Nevertheless, panoramic imaging presents important methodological limitations. Inherent magnification and geometric distortion may compromise the reliability of linear measurements and the accurate visualization of the alveolar crest level, especially when patient positioning is suboptimal [16,17]. In addition, its spatial resolution is inferior to that of intraoral radiographs, which may result in underestimation of early bone defects and limited characterization of critical areas such as interproximal crests and furcation involvements [18].

For these reasons, findings derived from panoramic radiography should be interpreted as orientative information rather than a substitute for intraoral radiographs when diagnostic precision or detailed longitudinal follow-up is required [2].

2.2.3. Periapical and Bitewing Radiographs

Periapical and bitewing radiographs represent the most valuable intraoral two-dimensional imaging modalities for detailed periodontal assessment, owing to their higher spatial resolution and the possibility of standardization through the paralleling technique and the use of positioning devices [11,19]. These images allow more accurate documentation of the interproximal alveolar crest level and facilitate correlation between radiographic findings and clinical parameters, which is essential for the correct application of the staging and grading system introduced in the 2017 Classification [5].

Bitewing radiographs are particularly useful in posterior regions, where they enhance visualization of interproximal bone levels and aid in detecting local contributing factors such as calculus deposits and subgingival restorative margins. Periapical radiographs provide complementary information on root morphology, lamina dura integrity, and periodontal ligament space [3,11].

Despite these advantages, both modalities share the inherent limitations of two-dimensional imaging, including the inability to assess the buccolingual dimension of osseous defects and the requirement for substantial mineral loss before radiographic changes become evident [20]. Consequently, their diagnostic value is maximized when integrated with comprehensive clinical examination and interpreted within a standardized, correlational framework [6].

2.3. World Workshop Classification 2017: Staging and Grading

The 2017 World Workshop classification of periodontal and peri-implant diseases introduced a multidimensional diagnostic framework that integrates disease severity, clinical management complexity, and the estimated risk of disease progression. This system was designed to overcome the limitations of previous classifications based on rigid clinical categories, offering a conceptual model more closely aligned with current biological and clinical evidence [4].

The staging component describes disease severity and treatment complexity at the time of diagnosis. Severity is primarily determined by interproximal clinical attachment loss, radiographic bone loss, and tooth loss attributable to periodontitis. Complexity criteria include deep periodontal pockets (≥6 mm), vertical bone defects, furcation involvement, tooth mobility, and functional disturbances that may influence therapeutic planning [4,5].

Clinical application of the staging and grading system has been described in observational studies that applied the 2017 Classification to characterize periodontal patient populations. Motegi et al. implemented the staging and grading framework in a clinical cohort. They reported the distribution of patients according to stage and grade, identifying variability among tooth types without proposing modifications to the original diagnostic criteria [21]. Similarly, Ertaş et al. applied the 2017 criteria for stage and grade determination in patients with periodontitis, demonstrating clinical feasibility without introducing additional diagnostic parameters [22].

The grading component introduces a dimension aimed at estimating the rate of disease progression. Grade assignment is primarily based on the ratio between radiographic bone loss and patient age as an indirect indicator of progression, adjusted by well-established risk modifiers such as smoking and diabetes mellitus [4]. From a conceptual perspective, narrative reviews have discussed the potential prognostic value of grading as well as challenges related to its clinical application, without providing original longitudinal data or independent prognostic validation [23].

Reproducibility of the staging and grading system has been assessed through studies evaluating diagnostic consistency and accuracy using standardized clinical cases. Marini et al. examined the accuracy and consistency of the 2018 system by comparing the performance of experts, general dentists, and undergraduate students, demonstrating that the classification can be applied consistently under controlled conditions. However, differences related to level of training were observed [24]. In a complementary study, Oh et al. analyzed inter-examiner agreement in stage and grade assignment using the 2017 Classification, identifying discrepancies among observers without extrapolating these findings to longitudinal clinical scenarios [25].

The available evidence indicates that the 2017 Classification represents a conceptually robust and clinically applicable diagnostic framework. Its use has been documented in clinical populations, evaluated for diagnostic consistency under standardized conditions, and analyzed regarding its adoption in academic settings [26]. However, current studies do not yet allow definitive conclusions regarding its longitudinal prognostic performance or its homogeneous implementation in real-world clinical practice. This underscores the need for structured training strategies, diagnostic calibration, and decision-support tools to optimize its implementation in contemporary periodontal practice [9].

2.4. Conventional Periodontal Microbiological Diagnosis

2.4.1. Principles of Anaerobic Culture Methods

Conventional periodontal microbiological diagnosis was historically based on identifying cultivable microorganisms from the subgingival biofilm using anaerobic techniques. This approach dominated periodontal microbiology during the second half of the twentieth century and enabled the establishment of the first associations between specific bacterial species and periodontal tissue destruction. These findings contributed to the transition from an unspecific concept of periodontitis to a polymicrobial infectious disease model [27,28].

The classical procedure involved collecting subgingival samples using sterile paper points or curettes, which were then transported in reducing media to preserve anaerobic bacteria. The samples were subsequently cultured on enriched selective and non-selective media and incubated under strictly anaerobic conditions for 5 to 14 days. Bacterial colonies were identified based on phenotypic characteristics such as colony morphology, Gram staining, carbohydrate fermentation, and enzymatic activity, allowing quantification of colony-forming units and estimation of the composition of the subgingival microbiota [29].

The primary value of culture-based methods resided in their ability to isolate viable bacteria and correlate their presence with specific clinical manifestations, thereby establishing the concept of “specific polymicrobial infection” in periodontology. Despite their technical complexity and strict anaerobic requirements, culture techniques constituted the cornerstone of periodontal microbiological diagnosis for decades, both in research settings and in the management of complex clinical cases [27,30].

The systematic application of culture-based techniques enabled the identification of a group of bacterial species closely associated with periodontal tissue destruction. Among the most extensively studied are Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, and Tannerella forsythia, which have been consistently recognized as key periodontal pathogens [31,32].

The repeated isolation of these microorganisms in patients with periodontitis allowed the establishment of robust clinical correlations. It supported the development of the microbial complex model (blue, orange, and red complexes). These ecological groupings remain a conceptual reference in contemporary periodontal microbiology, providing a framework to understand the structured organization and pathogenic potential of the subgingival biofilm [32].

2.4.2. Clinical Applications and Limitations of Anaerobic Culture Methods

Anaerobic culture techniques were clinically relevant between the 1980s and early 2000s, particularly in the assessment of aggressive, refractory, or recurrent periodontitis. Their main clinical role was to confirm the presence of specific periodontal pathogens, guide the selection of systemic antimicrobial therapy, and monitor treatment response and bacterial recolonization following subgingival instrumentation [33].

Over time, their clinical utility declined due to several limitations, including prolonged incubation periods that delayed therapeutic decisions [28], loss of viability of strict anaerobes during sample transport with consequent false-negative results [31], and limited sensitivity, as only 40–60% of the subgingival microbiome is cultivable [34]. Additional constraints include technical complexity, high costs, and the inability to capture the ecological interactions characteristic of the periodontal biofilm. Moreover, culture results reflect bacterial presence but do not necessarily indicate virulence activity or current host inflammatory status [34].

Nevertheless, anaerobic culture methods remain relevant in basic research and for phenotypic validation of bacterial strains, particularly in studies of antimicrobial resistance and as a reference for the development of subsequent molecular diagnostic techniques.

2.5. Targeted Molecular Diagnostic Tests

The development of molecular diagnostic techniques represented a major turning point in periodontal microbiology by enabling the accurate identification of non-cultivable bacteria and overcoming several limitations inherent to anaerobic culture-based methods. Since the 1990s, the introduction of polymerase chain reaction (PCR) assays and DNA–DNA hybridization systems has allowed detailed characterization of the subgingival biofilm with substantially higher sensitivity and specificity [34].

These methodologies marked a transition from predominantly phenotypic diagnostics toward a genotypic approach based on the detection of pathogen-specific genetic sequences. As a result, understanding of the subgingival microbiome expanded considerably, consolidating the concept of periodontal dysbiosis, in which pathogenicity does not rely on a single bacterial species but rather on the dynamic interaction of microbial communities with differential inflammatory potential [34].

2.5.1. Targeted Molecular Tests

• Polymerase Chain Reaction (PCR)

PCR transformed bacterial detection in periodontology by enabling amplification of specific microbial DNA sequences from minimal genetic material, without the need for prior cultivation [35]. This technique employs primers complementary to conserved regions of the 16S rRNA gene, present in most bacteria, or to pathogen-specific sequences targeting periodontal species such as Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, and Tannerella forsythia [36].

Technically, PCR relies on exponential amplification of bacterial DNA through controlled thermal cycles, generating millions of copies of the target region. Contemporary variants, including quantitative real-time PCR (qPCR), allow not only qualitative detection but also relative quantification of subgingival bacterial load, providing additional information on infection severity and potential risk of periodontal disease progression [34].

PCR exhibits high sensitivity and specificity, facilitating simultaneous detection of multiple bacterial species, including non-cultivable organisms. This capability has enabled the characterization of complex microbial profiles and their association with different stages of periodontitis [37]. Nevertheless, relevant limitations persist, such as detection of residual DNA from non-viable bacteria, the requirement for specialized equipment, and dependence on appropriate primer selection to avoid false-positive or false-negative results [38].

• DNA–DNA Hybridization

Before the widespread adoption of PCR, DNA–DNA hybridization methods represented the first genotypic approach for identifying periodontal bacteria without relying on conventional culture techniques [27]. In this method, microbial DNA fragments extracted from clinical samples are immobilized on a membrane and exposed to labeled probes complementary to specific 16S rRNA gene sequences. Probe–target binding is detected using colorimetric or fluorescent signals [34]. The classical checkerboard DNA–DNA hybridization technique enabled simultaneous analysis of more than 40 bacterial species across multiple subgingival samples. This approach provided a comprehensive overview of the periodontal microbiota. It was instrumental in defining microbial complexes (red, orange, green, and yellow), which remain a reference for the ecological organization of the periodontal biofilm.

Its main advantage lies in its multiparametric capacity to analyze numerous species within a single assay. However, sensitivity is lower than that of PCR, interpretation depends on DNA quantity and quality, and probe design is limited to previously characterized species, restricting detection of emerging or uncharacterized microorganisms [39].

• BANA Test (Benzoyl-DL-Arginine-Naphthylamide)

The BANA test is a rapid diagnostic tool based on indirect detection of enzymatic activity associated with periodontopathogenic bacteria. It relies on hydrolysis of the substrate benzoyl-DL-arginine-naphthylamide by trypsin-like proteolytic enzymes produced by P. gingivalis, T. forsythia, and Treponema denticola. Substrate degradation releases a colored compound, indicating a positive reaction. The BANA test has been used as a low-cost, rapid, and semi-quantitative screening tool, particularly in clinical settings with limited access to advanced molecular diagnostics [40].

However, its diagnostic utility is constrained by lower sensitivity and specificity compared with PCR or DNA–DNA hybridization. Moreover, it does not allow discrimination between individual species or quantification of microbial load and may yield positive reactions in the presence of other proteolytic bacteria with similar enzymatic activity [41].

2.5.2. Clinical Applications, Advantages, and Limitations

Targeted molecular tests have represented a significant advance in risk stratification and therapeutic monitoring of periodontitis. Their high sensitivity has enabled confirmation of specific periodontal pathogens, particularly in patients with atypical clinical presentations, and has facilitated correlations between microbial profiles and different disease stages and patterns [34,42].

In clinical practice, PCR is primarily used for targeted detection of Aggregatibacter actinomycetemcomitans and Porphyromonas gingivalis. In contrast, DNA–DNA hybridization and the BANA test are more frequently applied in screening or post-therapy follow-up settings, where a rapid and indicative assessment is required. The main advantages of these techniques include independence from bacterial growth, the ability to detect both cultivable and non-cultivable microorganisms, and shorter turnaround times compared with conventional culture-based methods [43].

Nevertheless, relevant clinical limitations persist, including higher costs, the need for specialized equipment, and difficulty distinguishing between active infection and detection of residual genetic material. Consequently, molecular test results should always be interpreted in conjunction with clinical and radiographic findings, avoiding therapeutic decisions based exclusively on isolated microbiological data [44].

Targeted molecular diagnostics have progressively displaced culture-based methods as the central diagnostic tool in research and in complex clinical cases, by providing a broader and more detailed characterization of the subgingival ecosystem and its role in periodontal pathogenesis.

2.6. Omics and Susceptibility Biomarkers

The evolution of periodontal diagnosis has shifted from conventional microbiological methods, primarily focused on culture-based techniques and targeted pathogen detection, toward an integrative approach with high biological resolution. Advances in molecular biology technologies and computational analysis have enabled comprehensive characterization of the periodontal ecosystem across multiple biological layers, incorporating genomic, transcriptomic, proteomic, and epigenomic data. This multi-omics framework has contributed to redefining periodontitis not as a localized infection, but as a complex inflammatory disorder driven by dynamic interactions between the host and the oral microbiome [45].

2.6.1. Clinical Applications, Advantages, and Limitations

The adoption of omics technologies has consolidated the understanding of periodontitis as the outcome of ecological dysbiosis and an altered host response, involving complex networks of genetic and metabolic regulation. Within this paradigm, pathogenicity is no longer attributed exclusively to individual bacterial species, but rather to the synergy between microbial communities and host-related determinants—including epigenetic and proteomic components—that modulate inflammation and tissue repair processes [46].

Emphasis has been focused on the fact that multi-omics integration combines information from different biological layers—DNA, RNA, proteins, and metabolites—through analytical strategies such as cross-correlation and co-expression network analysis, to identify functional relationships between the microbiota and the immune response. In periodontology, this convergence has facilitated the development of susceptibility bioprofiles and predictive models of disease progression, paving the way toward precision periodontology [47].

2.6.2. Clinical Applications, Advantages, and Limitations

The analysis of the subgingival microbiome using 16S rRNA gene sequencing and next-generation metagenomics has reduced reliance on culture-based methods, enabling the detection of both cultivable and non-cultivable microorganisms and allowing a more comprehensive characterization of the ecological balance of the subgingival biofilm [48].

A meta-analysis including more than 3,000 subgingival plaque samples reported that microbiome-derived profiles achieved a diagnostic accuracy (AUC) of 0.87 (95% CI: 0.83–0.90) for discriminating between periodontal health and disease. Taxa with the highest predictive value included Porphyromonas gingivalis, Tannerella forsythia, Treponema denticola, Filifactor alocis, and Fretibacterium fastidiosum, supporting the concept of an “expanded red complex” under metagenomic approaches [48].

Additionally, Li et al. described the application of machine learning algorithms to 16S rRNA datasets to predict periodontitis stage with reported diagnostic accuracy exceeding 90%, reinforcing the concept of the microbiome as a community-level biomarker—rather than an individual species-based marker—of disease activity [49].

From a functional perspective, metatranscriptomic profiling has enabled the identification of active bacterial metabolic pathways, including those associated with collagen degradation, lipopolysaccharide (LPS) production, and activation of host Toll-like receptor signaling. This approach integrates bacterial ecology with host inflammatory responses, providing a functional bridge between microbial composition and disease pathophysiology [44,50].

2.6.3. Transcriptomics and Proteomics: Expression of the Host Inflammatory Response

Transcriptomic analyses of gingival tissues and gingival crevicular fluid (GCF) have enabled the characterization of differential gene expression patterns in response to microbial challenge. Using microarray platforms and RNA sequencing (RNA-Seq), studies have reported overexpression of pro-inflammatory genes (e.g., IL1B, TNF, CXCL8, MMP9, TLR2) and activation of pathways related to extracellular matrix degradation and oxidative stress [51]. In parallel, proteomic approaches have facilitated the identification of specific proteins in saliva and GCF that reflect active periodontal tissue destruction. Among the most extensively characterized biomarkers are MMP-8, MMP-9, calprotectin, osteoprotegerin (OPG), IL-1β, IL-6, and TNF-α [52].

The validation of a panel of 13 molecular biomarkers in saliva and GCF with a diagnostic accuracy of 94% supports the use of biomarker panels rather than isolated markers for periodontal diagnosis [53]. Consistently, a meta-analysis that used multibiomarker combinations achieved superior diagnostic performance compared with single biomarkers, reinforcing the clinical relevance of multivariate proteomic panels for differentiating active inflammation, remission, and risk of recurrence in periodontitis [54].

2.6.4. Epigenomics: DNA Methylation and microRNA-Mediated Regulation

Epigenomics provides a framework for understanding mechanisms associated with individual susceptibility to periodontitis. DNA methylation and microRNA (miRNA)-mediated regulation modulate gene expression without altering the genomic sequence and reflect interactions among environmental exposures, microbial signaling, and host immune responses [55].

Distinct epigenetic patterns have been described in periodontal disease, including hypermethylation of promoters of anti-inflammatory genes (e.g., IL10, SOCS1) and hypomethylation of pro-inflammatory genes (e.g., IL6, TNF, MMP9, TLR4), which correlate clinically with probing depth and alveolar bone loss [56]. Importantly, these epigenetic signatures may partially normalize following periodontal therapy, suggesting a potential role as dynamic biomarkers of disease activity and therapeutic response [57].

In parallel, several miRNAs—particularly miR-146a, miR-155, and miR-21—have been reported to be overexpressed in gingival tissues and saliva from patients with periodontitis, with functional implications in the regulation of NF-κB signaling, cytokine production, and osteoclast differentiation. Their relative stability in biological fluids and detectability through quantitative PCR position miRNAs as promising candidates for non-invasive periodontal diagnostics [58].

2.6.5. Multi-Omics Integration and Predictive Modeling

The integration of multiple omics layers underpins the concept of precision periodontology. Integrative models have been proposed based on cross-correlations among gene expression profiles, proteomic signatures, and microbial composition to construct host–microbiome interaction networks. These frameworks enable the development of predictive algorithms that combine molecular biomarkers with conventional clinical parameters, thereby enhancing diagnostic resolution and risk stratification [59]. Multi-omics approaches provide superior predictive power and reproducibility by reducing the inherent bias of single-layer measurements. The integration of microbiome, transcriptomic, and epigenomic data through machine learning techniques has emerged as a promising strategy to anticipate disease progression and monitor therapeutic response at the individual level [60].

2.6.6. Advantages and Limitations

Among the main advantages of omics-based approaches are their high sensitivity, the ability to detect molecular alterations before overt clinical manifestations, and the integration of multiple biological layers into unified analytical models. This strategy provides a holistic view of the periodontal inflammatory process and opens new avenues for personalized diagnostic and therapeutic strategies. Nevertheless, several limitations remain. These include heterogeneity across analytical platforms, the lack of clinically validated cutoff values, high operational costs, and the requirement for advanced bioinformatics expertise to ensure robust data interpretation. In addition, many available studies remain exploratory, with limited longitudinal evidence to support large-scale prognostic validation [61].

Multi-omics integration combined with artificial intelligence-driven algorithms delineates a future diagnostic paradigm in which periodontal assessment is complemented by personalized molecular signatures reflecting host–microbiome interactions. Table 1 summarizes the principal omics platforms currently applied in periodontal research, highlighting their methodological basis, representative biomarkers, and level of clinical validation.

3.1. From Molecular Biology to Digital Periodontology

The diagnostic paradigm in periodontology is shifting from traditional clinical observation toward a quantitative, reproducible, and digitally driven characterization of periodontal damage. This transition is supported by the comprehensive digitalization of dental practice, where the integration of three-dimensional imaging, AI-based algorithms, and clinical–molecular data provides a more comprehensive view of the pathological process and its interindividual heterogeneity [9].

The incorporation of technologies such as cone-beam computed tomography (CBCT), intraoral scanners, and predictive models based on deep learning has enabled more precise quantification of alveolar bone loss and soft tissue involvement, thereby reducing diagnostic subjectivity. Farina et al. (2025) emphasize that the integration between radiographic information and intelligent algorithms marks the emergence of a digital periodontology era oriented toward personalized treatment strategies and dynamic disease monitoring [62].

Jacobs et al. further complement this perspective by highlighting that three-dimensional visualization of the alveolar bone facilitates the correlation of morphological findings with clinical variables and biological markers, fundamentally transforming the assessment of disease progression and therapeutic response. In this sense, contemporary periodontology is positioned at the intersection of biological insight and digital innovation [11].

3.2. Cone-Beam Computed Tomography and Volumetric Analysis

Cone-beam computed tomography (CBCT) is currently regarded as a reference standard for the three-dimensional assessment of periodontal defects in selected clinical scenarios. It provides voxel sizes ranging from 0.08 to 0.2 mm, enabling the detection of defects as small as 1 mm with an accuracy exceeding 90%. The technique is based on volumetric reconstruction from a cone-shaped X-ray beam, generating isotropic images across the three spatial planes [11].

CBCT allows comprehensive visualization of the alveolar bone and root morphology, as well as the assessment of their spatial relationship with critical anatomical structures, including the mandibular canal, maxillary sinus, and vestibular cortical plate. CBCT significantly improves detection of infrabony and furcation defects compared with periapical radiography, achieving 92% concordance with direct surgical observations. The main advantages of CBCT include the ability to obtain accurate three-dimensional measurements without geometric distortion, detailed visualization of bone phenotype and alveolar cortical plates, and the facilitation of digital planning for regenerative and mucogingival surgical procedures [62]. Despite these benefits, routine CBCT use is constrained by radiation dose, acquisition costs, and the potential presence of metallic artifacts. Accordingly, selective use is recommended, particularly in complex cases, combined defects, or for implant-related risk assessment [9].

Volumetric analysis using dedicated software platforms such as Amira, Dolphin, Mimics, or 3D Slicer enables the quantification of bone loss volume in cubic millimeters. It facilitates longitudinal comparisons—with a strong correlation between CBCT-derived measurements and histological assessments, supporting the reliability of this approach [11]. Furthermore, when combined with artificial intelligence, automated segmentation of osseous structures has substantially reduced analysis time, decreasing from approximately 45 minutes to less than 10 minutes [62].

3.3. Three-Dimensional Analysis and Volumetric Reconstruction

Three-dimensional analysis has transformed periodontal imaging into a quantitative diagnostic tool. The fusion of DICOM files derived from cone-beam computed tomography (CBCT) with STL files obtained from intraoral scanners enables the generation of hybrid models that integrate osseous and gingival information, allowing detailed three-dimensional visualization of the spatial relationship between the dental crown, alveolar bone, and gingival margin [63].

These hybrid models enable precise assessment of the topography of infrabony defects and dehiscences, facilitate longitudinal monitoring of post-surgical bone regeneration, and allow the correlation of tissue changes with molecular biomarkers or clinical parameters. Despite its considerable potential, methodological limitations persist, including the lack of standardized calibration protocols and heterogeneity among segmentation algorithms. Interobserver reproducibility may vary by up to 12% depending on the density threshold applied, underscoring the need for international consensus on image processing standards in periodontal diagnostics [11].

3.4. Artificial Intelligence in Periodontal Imaging

Artificial intelligence (AI) represents one of the most significant technological innovations in contemporary periodontal diagnosis. Deep learning–based systems, particularly convolutional neural networks (CNNs), have demonstrated a high capacity to detect, segment, and classify alveolar bone loss in panoramic and periapical radiographs [64].

Chang et al. developed a model that achieved an overall average accuracy of 0.87 ± 0.01 in categorizing mild (< 15%) or severe (≥ 15%) bone loss with fivefold cross-validation in identifying alveolar bone loss. These findings suggest clinically meaningful diagnostic equivalence and support the potential of AI systems for routine classification tasks in periodontal assessment [65]

Three-dimensional AI approaches, including 3D-CNNs and Vision Transformers applied to CBCT data, enable automatic segmentation of infrabony defects with area under the curve (AUC) values exceeding 0.95. When combined with clinical and demographic variables, these models can generate personalized risk maps, thereby supporting individualized therapeutic decision-making. The most relevant advantages of AI-based diagnostic systems include reduced interobserver variability, the ability to process large datasets, and increased diagnostic efficiency. Nevertheless, significant challenges remain, including the limited availability of multicenter datasets, potential population bias, and the need for robust ethical and regulatory frameworks to ensure safe clinical implementation [66].

3.5. Artificial Intelligence –Assisted Thermal Imaging

An emerging application within digital periodontal diagnosis is the use of artificial intelligence–assisted thermal imaging to assess gingival inflammation. Infrared thermography measures temperature gradients associated with inflammatory processes, as vasodilation and increased blood flow lead to elevated surface tissue temperature. An AI-based model using XGBoost and Random Forest algorithms was used to analyze gingival thermal patterns obtained with high-resolution infrared cameras. The algorithm achieved an accuracy of 92.7%, a precision of 92.8%, and an F1-score of 0.9278 in classifying degrees of gingival inflammation (mild, moderate, and severe), demonstrating a clear correlation between thermal intensity and clinical gingival index parameters [67]. This approach provides a non-invasive, radiation-free diagnostic modality with the potential to detect subclinical inflammation before overt clinical manifestation. AI-generated thermal maps can identify localized inflammatory microzones and quantify their extent, enabling longitudinal monitoring of treatment response.

Key advantages include applicability without ionizing radiation or direct tissue contact and improved diagnostic consistency by reducing reliance on subjective interpretation. Thermography may complement conventional radiographic methods and serve as a periodontal screening tool, particularly in primary care settings or teleperiodontology [9,62].

Nevertheless, these findings should be interpreted with caution, as outcomes may be influenced by environmental conditions, including ambient temperature and humidity, as well as individual physiological variability. Accordingly, rigorous standardization of thermal imaging protocols and camera calibration is essential to ensure reproducibility and facilitate broader clinical implementation. AI-assisted thermography provides a functional dimension to periodontal diagnostics by enabling the detection of inflammatory activity, thereby complementing conventional radiographic approaches that predominantly assess structural tissue alterations. This integrative perspective supports the development of multimodal predictive models that incorporate metabolic, thermal, and morphological parameters, offering a more comprehensive framework for disease characterization and risk stratification.

3.6. Integrated Digital Diagnosis (CBCT + AI + Intraoral Scanning)

Integrated digital diagnosis represents the culmination of the current technological paradigm in periodontology. It is based on the fusion of anatomical, morphological, and functional three-dimensional models to construct a periodontal “digital twin,” defined as a computational representation of the patient that combines CBCT imaging, optical records, and predictive algorithms [62].

The interoperability between DICOM and STL files enables comprehensive reconstructions of the dentogingival complex, which are valuable for surgical planning and longitudinal follow-up. This approach facilitates the design of personalized digital guides, preoperative simulations, and long-term tissue stability analysis. Integrated digital models have also demonstrated utility in ortho-periodontics and reconstructive surgery by allowing precise three-dimensional assessment of tooth position and soft tissues in relation to the alveolar bone. A strong correlation has been reported between clinical measurements and integrated digital models, supporting their applicability in both research and advanced clinical practice [68].

From a conceptual perspective, digital integration aligns with the 2018 periodontal classification and the EFP S3 clinical guidelines by providing a more objective framework for documenting disease severity, treatment complexity, and therapeutic response. The convergence of imaging technologies, artificial intelligence, and clinical data may foster the development of a predictive and hybrid form of periodontology, in which patients are monitored through digital models integrating biomarkers, bone topography, and functional inflammatory parameters. Remaining challenges include interplatform standardization, biomedical data protection, and the need for clinician training in advanced digital interpretation. Nevertheless, integrated digital diagnosis is emerging as a key component of periodontal management in the coming decade.

3.7. Biological Fluids and Periodontal Biomarkers

Periodontal diagnosis has evolved from static clinical parameters toward a dynamic model aimed at detecting biological activity and estimating the risk of disease progression. In this context, biological fluids—primarily saliva, gingival crevicular fluid, and to a lesser extent peripheral serum—have emerged as non-invasive diagnostic matrices that reflect local inflammatory and tissue status. These analyses enable the quantification of inflammatory mediators, proteolytic enzymes, bone-related molecules, and genetic or epigenetic profiles that characterize host–microbiome interactions [69].

Over recent years, periodontal biomarker research has shifted from a “single-marker” approach toward multivariate panels integrating different biological pathways, including cytokines, matrix metalloproteinases, microRNAs, and bacterial DNA. This transition reflects the multifactorial nature of periodontitis and the need for complementary tools that enhance clinical and radiographic assessment, improving risk stratification and therapeutic monitoring [34,70].

3.7.1. Saliva as a Diagnostic Matrix

Saliva represents an accessible and stable source of periodontal biomarkers. It consists of secretions from major and minor salivary glands, GCF exudate, epithelial cells, bacteria, and metabolites. Its composition reflects both the local oral environment and systemic host responses, making it a valuable tool for global monitoring of inflammatory processes [69].

o

Inflammatory Mediators and Matrix Metalloproteinases

Pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) and matrix metalloproteinases (MMP-8 and MMP-9) are among the most extensively studied salivary biomarkers. MMP-8, also known as collagenase-2, is secreted by activated neutrophils and is directly associated with gingival collagen degradation and clinical attachment loss. Its levels are significantly elevated in active periodontitis and decrease following mechanical therapy, correlating with clinical improvement. MMP-9 and calprotectin (S100A8/A9) complement this profile by reflecting neutrophil-driven inflammation and connective tissue activity. In a meta-analysis, Blanco-Pintos et al. (2023) demonstrated that the combination of MMP-8, IL-1β, and IL-6 improved diagnostic accuracy (AUC = 0.88) compared with single biomarkers, supporting the use of proteomic panels to discriminate between periodontal health, gingivitis, and active periodontitis with sensitivities exceeding 85%[70].

o

MMP-8 Point-of-Care Tests

The development of rapid tests for active MMP-8 (aMMP-8) represents a significant step toward clinical translation. These point-of-care tests (POCTs) allow detection in mouth rinses or whole saliva within less than 10 minutes. Multicenter studies have reported diagnostic accuracies of 80–90% for identifying patients with periodontitis [71]. Their main advantages include portability and ease of use for population screening and post-treatment monitoring. However, interindividual variability and the need for standardized reference values remain relevant challenges.

o

Microbial and Genetic Signatures in Saliva

Saliva also serves as a matrix for detecting microbiome-related biomarkers. Analyses based on 16S rRNA sequencing and quantitative PCR have identified specific bacterial profiles, with the presence of Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola correlating with advanced stages of periodontitis (Na et al., 2020). In addition, increased relative abundance of Filifactor alocis and Prevotella intermedia has been associated with persistent and refractory inflammation [34].

At the epigenetic level, salivary microRNAs have emerged as promising biomarkers. In particular, miR-146a and miR-155 are overexpressed in patients with active periodontitis and are linked to activation of the NF-κB pathway and cytokine regulation [72]. These molecules may serve as prognostic biomarkers and potentially link periodontitis to systemic comorbidities such as cardiovascular disease and type 2 diabetes.

Saliva collection is simple, non-invasive, and reproducible, making it suitable for both clinical and community-based applications. It is compatible with multiplex technologies, including ELISA, Luminex assays, and microfluidic biosensors. However, salivary biomarker levels represent a heterogeneous mixture of sites and can be influenced by salivary flow, circadian rhythms, diet, smoking, and bleeding. Therefore, salivary results should be interpreted in conjunction with clinical findings and, when feasible, complemented by GCF analysis [70,71].

3.7.2. Gingival Crevicular Fluid (GCF)

GCF represents a more site-specific matrix derived from the inflammatory exudate of the gingival sulcus or periodontal pocket. It contains cellular products, plasma proteins, inflammatory mediators, and tissue enzymes that closely reflect local destructive activity [71].

o

Inflammatory and Bone-Related Markers

Key GCF biomarkers include IL-1β, IL-6, TNF-α, MMP-8, MMP-9, calprotectin, and the RANKL/OPG axis. IL-1β and IL-6 correlate with probing depth and bleeding on probing, while MMP-8 and MMP-9 reflect collagenolytic activity and decrease following non-surgical therapy. Elevated RANKL levels combined with reduced OPG are associated with active alveolar bone resorption [53]. Cafiero et al. reported that the combination of MMP-8 and the RANKL/OPG ratio predicted disease progression with a sensitivity of 82% and specificity of 88%. Blanco-Pintos et al. further confirmed the utility of GCF protein panels for periodontal phenotype classification and therapeutic monitoring.

o

Dynamics and Clinical Utility

GCF responds rapidly to changes [71] in bacterial load and inflammatory status, making it particularly suitable for site-specific assessment of disease activity and treatment response. Unlike saliva, GCF can be anatomically correlated with individual periodontal sites, allowing detection of active lesions before clinically evident attachment loss. This characteristic supports its use in longitudinal research and periodontal regeneration trials [34].

o

Methodological Limitations

GCF collection requires standardized techniques (microcapillary tubes or paper strips), strict contamination control, and operator calibration. Variations in sampling time and anatomical site can influence results. In addition, heterogeneity among analytical methods (ELISA, Western blot, Luminex) limits interstudy comparability, underscoring the need for internationally harmonized protocols [71].

3.7.3. Systemic Biomarkers

Periodontitis is associated with systemic inflammatory responses that can be assessed using serum biomarkers such as C-reactive protein (CRP), IL-6, TNF-α, and adipokines. Correlations between CRP and IL-6 levels and Grade C of the 2017 Classification have been demonstrated, reinforcing the link between periodontal disease and cardiovascular risk [70].

More recently, integrating local and systemic biomarkers using machine learning approaches has enabled the development of predictive models for periodontal risk. Dong et al. analyzed multi-omics datasets and showed that combining microbiome, proteomic, and microRNA data achieved an AUC of 0.93 for discriminating periodontal health from disease. These algorithms can incorporate demographic, genetic, and clinical variables, supporting personalized risk stratification and prediction of therapeutic response [48].

Biological fluid–derived biomarkers provide complementary information to clinical and radiographic parameters, allowing a more comprehensive characterization of periodontal status. Nevertheless, their clinical implementation remains challenged by methodological heterogeneity and the lack of universally validated diagnostic thresholds [70,71]. While integration with artificial intelligence and digital platforms offers promising perspectives for automated diagnostic systems based on saliva or GCF, widespread adoption depends on longitudinal validation, cost-effectiveness, and demonstration of tangible clinical benefit.

From a translational standpoint, integrating molecular biomarkers, genetic information, and advanced imaging may enable the future development of periodontal “digital twins” capable of real-time disease monitoring. Table 2 presents emerging diagnostic technologies that complement conventional clinical and radiographic assessment in periodontal practice.

Summary of novel diagnostic approaches for periodontitis, highlighting their primary biological targets and key advantages. These technologies complement conventional clinical and radiographic assessment by enabling earlier detection, improved risk stratification, and greater diagnostic precision, supporting the transition toward precision periodontology.

4.1. Context and Evolution

Contemporary periodontology has progressively shifted from predominantly descriptive diagnostic models—based on clinical and radiographic parameters—toward a predictive and preventive framework driven by the integration of biomarkers, digital data, and artificial intelligence (AI). Traditionally, periodontal risk assessment relied on multivariable linear models, such as logistic regression and discriminant analysis, incorporating clinical variables (probing depth, clinical attachment loss, bleeding on probing, plaque levels) together with modifying factors (smoking status, age, systemic comorbidities) [73]. Although these approaches enabled basic risk stratification, they were constrained by assumptions of linearity, limited capacity to model complex interactions, and poor scalability for large, high-dimensional molecular or imaging datasets.

Advances in bioinformatics, coupled with the availability of omics databases (transcriptomics, microbiomics, epigenomics) and the digitalization of dental imaging (CBCT, panoramic and periapical radiographs), have facilitated the adoption of machine learning (ML) and deep learning (DL) algorithms capable of identifying patterns beyond the reach of conventional statistical methods. This convergence has contributed to the emergence of a predictive paradigm in periodontology, in which AI is not restricted to classification tasks but is increasingly oriented toward anticipating disease progression and therapeutic response through the integration of clinical, biological, and digital data streams [9].

4.2. Context and Evolution

Predictive models applied to periodontology can be broadly categorized as follows:

4.2.1. Classical Statistical Models

Multivariable logistic regression and Cox proportional hazards models estimate disease presence or progression risk based on predefined input variables [50]. While these models are relatively transparent and interpretable, their ability to capture non-linear relationships and complex interdependencies is limited. Furthermore, they require prior variable selection and show reduced performance when applied to large, heterogeneous datasets.

4.2.2. Machine Learning Models

Machine learning approaches (including Random Forest, Support Vector Machines, Gradient Boosting, and Artificial Neural Networks) identify data-driven patterns without assuming linear relationships among variables. In periodontology, these models have been applied to classify patients according to disease severity and grading, predict disease progression by integrating clinical and molecular features, and combine radiographic, clinical, and salivary information into hybrid predictive frameworks that enhance diagnostic and prognostic performance [48].

4.2.3. Deep Learning

DL, particularly convolutional neural networks (CNNs), has revolutionized automated interpretation of radiographic images. CNN-based models can segment alveolar bone and detect bone loss with accuracy comparable to, or exceeding, that of expert examiners, enabling automated severity grading and recognition of characteristic bone loss patterns in panoramic and periapical radiographs [65].

4.2.4. Input Variables and Multimodal Approach

Predictive models in periodontology integrate data across multiple domains, including:

o

Clinical variables: probing depth, clinical attachment loss, bleeding on probing, tooth mobility.

o

Demographic and behavioral factors: age, sex, smoking status, diabetes, oral hygiene habits.

o

Molecular markers: MMP-8, IL-1β, IL-6, RANKL/OPG, microRNA profiles, DNA methylation signatures.

o

Microbiome features: relative abundance of Porphyromonas gingivalis, Tannerella forsythia, Treponema denticola, Filifactor alocis.

o

Imaging-derived metrics: bone texture, alveolar crest loss, cortical density, and volumetric parameters extracted from CBCT datasets.

This heterogeneity underscores the need for multimodal predictive models capable of integrating dynamic, high-dimensional data of diverse biological and clinical origin, forming the basis for individualized risk assessment and precision periodontal care.

4.3. Clinical and Prognostic Applications

4.3.1. Automated Diagnosis and Staging Classification

Convolutional neural networks (CNNs) applied to panoramic radiographs have demonstrated a high capacity to detect alveolar bone loss, estimate the percentage of remaining bone, and approximate stage and grade assignment according to the 2018 periodontal classification. Chang et al. (2020) reported an accuracy of 89.3% for identifying alveolar bone loss using a CNN model trained on approximately 1000 panoramic radiographs [65].

In a complementary study, Farina et al. (2025) reported that integrating radiographic artificial intelligence with clinical variables significantly improves automated periodontitis classification, achieving a mean AUC of 0.93. These systems provide substantial advantages in diagnostic standardization and reduction of interobserver variability, particularly in high-volume clinical settings.

4.3.2. Prediction of Disease Progression and Therapeutic Response

Machine learning models based on longitudinal cohorts have enabled the identification of patients at increased risk of developing further clinical attachment loss or gingival recession following periodontal therapy. Bashir et al. compared six predictive algorithms and observed superior performance for Random Forest and Gradient Boosting models, with AUC values ranging from 0.87 to 0.91. Key predictors included baseline probing depth, bleeding on probing, diabetes mellitus, and salivary levels of MMP-8 and IL-6.

From a molecular perspective, Dong et al. integrated subgingival microbiome (16S rRNA) and salivary proteomic data, reporting an AUC of 0.93 for discriminating periodontal health from periodontitis, consistently outperforming models based exclusively on clinical variables. Together, these findings support the utility of machine learning in combining clinical and biological information into hybrid predictive frameworks.

4.3.3. Personalized Risk Models

A multivariable predictive model for severe periodontitis incorporating demographic, behavioral, clinical, and systemic factors established a basis for individualized risk assessment, despite showing moderate performance. Subsequent approaches enhanced this framework by integrating bacterial biomarkers, resulting in a notable improvement in predictive accuracy [74].

In a recent systematic review, Polizzi et al. reported that artificial intelligence–based models achieve a mean accuracy of 92% (95% CI: 88–96) for periodontitis prediction, consistently outperforming traditional statistical models. However, the authors emphasized that limited multicenter external validation remains a major barrier to clinical generalizability [75].

4.4. Performance and Comparative Evidence

The performance of predictive models in periodontology is commonly assessed using metrics such as sensitivity, specificity, accuracy, and F1-score. Deep learning models based on panoramic radiographs generally show more consistent and higher discriminative performance, whereas those using clinical or molecular data exhibit greater variability. Machine learning approaches like Random Forest outperform traditional logistic regression, particularly in modeling complex non-linear interactions, with hybrid models integrating multi-source data improving robustness and offering interpretations more aligned with periodontal pathophysiology [73,76].

4.5. Current Challenges and Future Perspectives

4.5.1. Interpretability and Transparency

Artificial intelligence models, particularly those based on deep learning, may function as “black boxes,” limiting their clinical adoption [65]. In this context, explainable AI (XAI) strategies are required to identify which variables or radiographic regions contribute most significantly to model predictions [75].

4.5.2. Population Bias and External Validation

A recurring limitation is that many models are trained on geographically restricted cohorts, increasing the risk of demographic bias. Farina et al. and Herrera et al. emphasize the need for multicenter studies with external validation and harmonized databases following FAIR principles (Findable, Accessible, Interoperable, Reusable).

4.5.3. Integration into Digital Clinical Practice

Current trends point toward integrated platforms combining artificial intelligence, CBCT, salivary biomarkers, and electronic health records, enabling “periodontal digital twin” approaches to simulate disease progression and estimate therapeutic response [9,73].

4.5.4. Ethical and Regulatory Considerations

The clinical implementation of AI systems raises challenges related to diagnostic responsibility, data protection, and algorithm reproducibility. Emerging regulatory frameworks for medical AI must therefore be explicitly considered in clinical adoption.

4.5.5. Education and Interoperability

Effective application of these models requires competencies in data analysis and critical interpretation of algorithmic outputs. Furthermore, interoperability with electronic health record standards, such as HL7 and FHIR, is essential for practical integration. The integration of predictive models, omics datasets, and advanced imaging technologies is likely to shape a more personalized and proactive approach to periodontal diagnosis and management.

5.1. Biologically and Risk-Based Periodontal Therapy

Modern periodontal therapy has progressively shifted from a predominantly mechanical and non-specific model toward biologically informed, mechanism-based, and risk-adapted strategies. Periodontitis is currently understood as a dysbiosis-driven inflammatory disease occurring in a susceptible host, in which bacteria are necessary but not sufficient to explain tissue destruction. Rather, the host inflammatory response to the microbial challenge primarily drives connective tissue degradation and alveolar bone loss [77].

Although nonsurgical periodontal therapy (scaling and root planing, with or without adjunctive antimicrobials) remains the standard of care for biofilm control, targeting microbes alone does not achieve optimal outcomes in all patients [77]. Interindividual variability in immune responsiveness, systemic comorbidities, genetic background, and environmental exposures contributes to different clinical trajectories despite similar microbial burdens. Consequently, modern periodontal diagnosis incorporates risk factors such as smoking, diabetes, and systemic inflammatory conditions, supporting a transition toward precision-based therapy.

5.2. From Non-Specific Mechanical Therapy to Mechanism-Driven Interventions

Historically, periodontal treatment was based on an infectious disease model focused primarily on plaque removal. However, advances in periodontal immunobiology have demonstrated that tissue breakdown is largely mediated by host-derived inflammatory pathways, including excessive cytokine production, matrix metalloproteinase (MMP) activation, and osteoclastogenic signaling [78,79].

Host-modulation therapy (HMT), developed nearly three decades ago, represents a pharmacologic strategy aimed at modifying the host’s destructive inflammatory and collagenolytic responses [78]. Subantimicrobial-dose doxycycline (SDD) remains the only FDA-approved host-modulation agent for periodontitis. Its mechanism includes inhibition of MMP activity and modulation of abnormal collagen degradation, demonstrating clinical efficacy as an adjunct to scaling and root planing [78,79].

Two principal categories of host-modulation therapy have been described:

▪

Modulation of the inflammatory response (through inhibition or resolution mechanisms).

▪

Modulation of pathologic collagenolytic activity within periodontal tissues [78].

Earlier attempts to reduce periodontal destruction through long-term administration of nonsteroidal anti-inflammatory drugs (NSAIDs) showed reductions in bone resorption. Still, they were ultimately discontinued due to adverse effects and rebound phenomena following drug cessation [80]. These findings reinforced the need for safer and more targeted biologic interventions.

5.3. Targeting Immune Pathways in Periodontal Therapy

Periodontitis shares inflammatory mechanisms with systemic conditions such as rheumatoid arthritis (RA), including dysregulation of TNF-α, IL-1β, IL-6, and RANKL pathways. Clinical observations indicate that patients receiving disease-modifying antirheumatic drugs (DMARDs) may exhibit improvements in periodontal parameters following immune modulation. For example, biologic therapies targeting CD20 (rituximab) and IL-6 receptors (tocilizumab) have been associated with reductions in periodontal inflammation [81]. These findings highlight shared inflammatory networks and open opportunities for mechanism-based therapeutic approaches.

Additional adjunctive strategies with anti-inflammatory potential include probiotics, omega-3 fatty acids, and statins. Probiotics may suppress NF-κB signaling and reduce proinflammatory cytokine production while enhancing regulatory T-cell responses. Omega-3 fatty acids and related lipid mediators contribute to pro-resolving pathways, supporting a shift from inflammation persistence to tissue homeostasis. Statins have also demonstrated anti-inflammatory and bone-preserving effects [82].

Conversely, some pharmacologic agents, such as bisphosphonates, have raised safety concerns due to the risk of medication-related osteonecrosis of the jaw, limiting their indication as adjunctive periodontal therapy [83].

5.4. Resolution of Inflammation as a Therapeutic Objective

A significant paradigm shift in periodontology is the recognition that resolution of inflammation is an active, highly regulated biological process. Specialized pro-resolving mediators (SPMs), including resolvins derived from omega-3 fatty acids, do not suppress acute inflammation but instead prevent its chronic persistence and promote tissue repair [84].

Preclinical studies have demonstrated that resolvin E1 reduces alveolar bone loss and promotes restoration of periodontal tissues in experimental models [85]. Emerging clinical data also suggest that adjunctive omega-3 supplementation, particularly when combined with low-dose aspirin, may enhance clinical outcomes and reduce inflammatory biomarkers following periodontal therapy [86].

Recent translational research highlights additional emerging targets, including complement inhibitors and other immune-modulatory agents currently under clinical investigation [87]. These developments reflect a conceptual transition from broad inflammatory suppression toward precision-based immune modulation and active orchestration of inflammatory resolution.

5.5. Integration of Biological Mechanisms and Risk-Based Care

A biologically and risk-based periodontal therapeutic model integrates:

▪

Mechanical biofilm control

▪

Targeted host-modulatory therapy

▪

Promotion of inflammatory resolution pathways

▪

Management of systemic and behavioral risk factors

By aligning diagnosis with underlying immunological mechanisms and patient-specific risk profiles, contemporary periodontal therapy advances toward precision medicine. The objective extends beyond microbial reduction to recalibration of the host response and long-term restoration of periodontal homeostasis.

5.6. Integration of Biological Mechanisms and Risk-Based Care

Contemporary periodontology has undergone a conceptual transformation from a morphology-based diagnostic paradigm toward a multidimensional framework integrating disease severity, biological activity, and patient-specific risk profiles. The 2017 World Workshop classification [5] introduced staging and grading as complementary dimensions to characterize both the extent of tissue destruction and the anticipated rate of progression [88]. However, staging and grading do not constitute prescriptive therapeutic algorithms. Rather, they serve as structured decision-support tools that inform the intensity, scope, and monitoring strategy of periodontal therapy within a personalized care model.

5.6.1. Staging, Grading, and Therapeutic Intensity

Staging primarily reflects the severity and complexity of periodontal breakdown, including clinical attachment loss, radiographic bone loss, probing depth, and functional impairment. Higher stages (III–IV) imply advanced tissue destruction and increased rehabilitative complexity [88]. From a therapeutic standpoint, staging informs the structural demands of care (such as the need for surgical intervention, regenerative procedures, or interdisciplinary rehabilitation) without dictating a rigid treatment pathway.

Grading, by contrast, provides an estimate of biological behavior and future risk. It incorporates direct or indirect evidence of progression and risk modifiers such as smoking and diabetes. Patients classified as Grade C demonstrate accelerated progression and heightened inflammatory responsiveness, suggesting the need for closer monitoring and potentially adjunctive systemic or host-modulatory strategies [47]. Importantly, grading shifts therapeutic reasoning from static assessment toward dynamic risk estimation. This approach supports calibrating treatment intensity and maintenance intervals according to individualized risk rather than uniform protocols.

5.6.2. Inflammatory Activity and Biomarkers as Indicators of Disease Dynamics

Traditional clinical parameters (probing depth, bleeding on probing, radiographic bone loss) primarily reflect cumulative tissue destruction. They provide limited insight into current inflammatory activity. Incorporating biomarkers offers the possibility of identifying ongoing biological processes and refining therapeutic decisions [89].

Systemic and local inflammatory markers—including IL-1β, IL-6, TNF-α, and C-reactive protein (CRP)—have been shown to correlate with periodontal severity and to decrease following non-surgical periodontal therapy [90]. Elevated salivary and GCF levels of IL-1β, RANKL, and IL-17 have been associated with advanced stages and higher grades of disease, supporting their role as indicators of inflammatory burden and progression risk [91].

Point-of-care testing technologies further enable real-time biomarker detection in saliva or GCF, allowing clinicians to monitor disease activity and therapeutic response more dynamically. For example, matrix metalloproteinase-8 (MMP-8) levels have been proposed as markers of active collagen breakdown and treatment response, facilitating individualized recall strategies and therapeutic adjustments [89].

Biomarker integration does not replace clinical judgment nor function as an independent therapeutic directive. Instead, biomarkers complement staging and grading by adding a biological dimension to clinical classification. Their greatest value lies in identifying subclinical activity, stratifying progression risk, and tailoring maintenance intensity.

5.6.3. Risk Stratification and Personalized Periodontal Care

Risk stratification represents the bridge between diagnosis and therapeutic personalization. Periodontitis progression is influenced by genetic susceptibility, systemic comorbidities, environmental exposures, and host immune responsiveness. Advances in genomics and multi-omics profiling have identified susceptibility loci (e.g., GLT6D1, FCER1G, HMCN2) associated with aggressive or severe phenotypes, highlighting the biological heterogeneity underlying similar clinical presentations [47].

Data-driven precision diagnostics integrate clinical parameters, biomarker profiles, imaging data, and molecular signatures to define biologically meaningful endotypes [47]. Machine learning and artificial intelligence models have demonstrated potential in synthesizing multidimensional data to predict disease trajectories and optimize individualized therapeutic strategies.

Staging and grading provide structural and prognostic orientation; inflammatory biomarkers inform current activity; and risk stratification integrates systemic and molecular determinants [79]. Together, these components support a graduated therapeutic model in which:

o

Low-risk, stable patients may require conventional mechanical therapy and standard maintenance intervals.

o

Moderate-risk individuals may benefit from adjunctive anti-inflammatory or host-modulatory strategies and closer follow-up.

o

High-risk or Grade C patients may require intensified surveillance, systemic risk control, and potentially personalized adjunctive interventions.

This stratified model avoids rigid therapeutic algorithms. Instead, it promotes adaptive clinical reasoning grounded in biological understanding and continuous reassessment.

5.6.4. Toward a Dynamic, Precision-Oriented Therapeutic Paradigm

The integration of staging, grading, inflammatory biomarkers, and molecular profiling reflects a paradigm shift from retrospective diagnosis toward predictive and preventive periodontology. While the current evidence supports the diagnostic and prognostic utility of these parameters, standardized biomarker-based therapeutic thresholds remain under development [47].

Modern periodontal decision-making should be understood as probabilistic and individualized rather than protocol-driven. Personalized periodontal therapy emerges not from isolated diagnostic categories but from the synthesis of structural damage assessment, biological activity evaluation, and patient-specific risk profiling.

5.7. Emerging Therapeutic Targets in Periodontitis

5.7.1. Inflammatory Mediators and Modulation of the Host Response

Periodontitis is no longer conceptualized solely as a biofilm-driven infection but as a dysregulated host inflammatory response to microbial dysbiosis. Advances in molecular immunology have identified specific cytokine networks and resolution pathways as central drivers of tissue destruction and potential therapeutic targets. Contemporary strategies therefore aim not only to reduce microbial burden but also to modulate host inflammatory cascades and promote resolution of inflammation.

5.7.2. Pro-Inflammatory Cytokines in Periodontal Disease

A central pathogenic axis in periodontitis involves the overexpression of pro-inflammatory cytokines such as interleukin-1 beta (IL-1β), tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and chemokines such as CCL2 (MCP-1). These mediators orchestrate leukocyte recruitment, osteoclastogenesis, matrix degradation, and amplification of inflammatory signaling.

IL-1β and TNF-α are considered master regulators of periodontal inflammation. They stimulate the production of matrix metalloproteinases (MMPs), prostaglandin E2 (PGE2), and receptor activator of nuclear factor kappa-B ligand (RANKL), thereby promoting connective tissue breakdown and alveolar bone resorption. Elevated levels of these cytokines have been consistently detected in gingival crevicular fluid (GCF), saliva, and serum of patients with periodontitis, correlating with disease severity [92].

IL-6 plays a dual role in immune regulation but contributes significantly to chronic periodontal inflammation by promoting pro-inflammatory T helper 17 (Th17) differentiation and enhancing osteoclast formation via RANKL expression. Increased IL-6 levels are associated with both local periodontal destruction and systemic inflammatory burden, supporting the concept of shared inflammatory networks between periodontitis and systemic diseases [93].

CCL2 (monocyte chemoattractant protein-1) is critical for monocyte/macrophage recruitment into periodontal tissues. By sustaining macrophage infiltration, CCL2 amplifies cytokine release and contributes to chronicity. Its upregulation in periodontal lesions highlights the importance of chemokine signaling as a therapeutic target aimed at disrupting inflammatory cell trafficking.

Among the Th17-related cytokines, IL-17 has emerged as a nodal mediator at the interface between microbial dysbiosis and host immunity [94]. IL-17 stimulates epithelial, stromal, and osteoblastic cells to release IL-1β, IL-6, TNF-α, chemokines, and MMPs, thereby amplifying inflammation and promoting osteoclastogenesis through RANKL induction [95]. Periodontal pathogens such as Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans enhance Th17 responses, sustaining a pro-resorptive microenvironment.

However, IL-17 biology is context-dependent. It also contributes to mucosal defense and neutrophil-mediated microbial control. This duality underscores the need for therapeutic strategies that modulate rather than completely suppress inflammatory pathways [92].

5.7.3. Resolution Mediators and Host Modulation Therapies

While traditional anti-inflammatory approaches aim to suppress inflammatory mediators, emerging paradigms emphasize the active resolution of inflammation. Resolution is a biologically regulated process mediated by specialized pro-resolving lipid mediators (SPMs) such as lipoxins, resolvins, protectins, and maresins. These molecules promote clearance of inflammatory cells, enhance efferocytosis, and restore tissue homeostasis without inducing immunosuppression.

Preclinical studies have demonstrated that resolvin E1 can reduce inflammatory infiltrates, inhibit osteoclastogenesis, and promote periodontal tissue regeneration. Rather than blocking upstream cytokines indiscriminately, resolution mediators redirect the inflammatory response toward homeostasis.

Host modulation therapy (HMT) represents a clinically established approach aligned with this concept. Sub-antimicrobial dose doxycycline is the most recognized example, functioning as an MMP inhibitor and modulator of collagen breakdown. By attenuating host-derived tissue destruction while preserving antimicrobial defense, SDD exemplifies mechanism-based adjunctive therapy [79].

Non-steroidal anti-inflammatory drugs (NSAIDs), anti-cytokine biologics (e.g., TNF inhibitors), and emerging IL-17-targeted therapies have been explored, particularly in patients with systemic inflammatory diseases. However, systemic cytokine blockade in periodontitis must balance potential benefits in reducing bone resorption against risks related to immune competence and mucosal defense [93].

Recent molecular insights suggest that therapeutic targeting of the IL-17/RANKL axis, chemokine signaling (e.g., CCL2), or downstream osteoclast pathways may offer more selective strategies. Additionally, modulation of macrophage polarization (M1 to M2 transition) and enhancement of regulatory T-cell responses are being investigated as biologically refined approaches.

5.7.4. Toward Mechanism-Guided Precision Therapy

The identification of cytokine-driven endotypes in periodontitis supports a transition from uniform mechanical therapy toward mechanism-guided adjunctive interventions [96]. Patients exhibiting high IL-17 or IL-6 signatures, exaggerated Th17 responses, or systemic inflammatory comorbidities may benefit from intensified host-modulatory strategies [97].

Importantly, emerging therapeutic targets do not replace conventional biofilm control but complement it. Mechanical debridement remains foundational, yet its biological effects—such as reduction in IL-17 levels following therapy—illustrate the dynamic interplay between microbial reduction and host response modulation [93].

Future research integrating multi-omics profiling, biomarker stratification, and clinical phenotyping may allow clinicians to identify patients with cytokine-dominant inflammatory profiles and tailor adjunctive host-targeted therapies accordingly. In this evolving paradigm, the goal is not indiscriminate immunosuppression but calibrated immunoregulation that restores periodontal homeostasis while preserving protective immunity.

5.8. Adaptive Immunity and the Th17/Treg Axis in Periodontitis

Adaptive immune responses are central to the transition from protective gingival immunity to chronic periodontal destruction. Among CD4⁺ T-cell subsets, the balance between Th17 cells and immunosuppressive regulatory T cells (Tregs) is now recognized as a pivotal axis controlling periodontal tissue homeostasis versus pathology. Under physiological conditions, Tregs—through IL-10, TGF-β, and cell–cell regulatory mechanisms—limit excessive inflammation and maintain tolerance to the oral microbiota. In periodontitis, however, this balance shifts toward a Th17-dominant response, promoting sustained inflammation and osteoclastogenic activity [98,99].

5.8.1. Pathogenic Role of Th17 Cells in Periodontal Destruction

Th17 cells, characterized by expression of RORγt and secretion of IL-17A, IL-17F, IL-21, and IL-22, play a dual role in mucosal immunity—protective against extracellular pathogens yet highly pathogenic when dysregulated. In periodontitis, elevated numbers of Th17 cells and increased levels of IL-17 and IL-23 have been consistently detected in gingival tissues, gingival crevicular fluid, and peripheral blood. These cytokines amplify local inflammation by stimulating fibroblasts, epithelial cells, macrophages, and osteoblast-lineage cells to produce IL-6, TNF-α, matrix metalloproteinases, and RANKL, thereby enhancing osteoclast differentiation and alveolar bone resorption [45,96,99].

Importantly, Th17 pathogenicity is closely linked to IL-23 signaling, which stabilizes the inflammatory Th17 phenotype and enhances tissue-destructive functions. An increased Th17/Treg ratio—rather than Th17 expansion alone—correlates strongly with disease severity, highlighting that insufficient regulatory control is as critical as excessive effector activity. Moreover, Th17 plasticity adds complexity: while many Th17 cells sustain IL-17 production in lesions, subsets may acquire alternative phenotypes that either exacerbate inflammation or, in specific contexts, contribute to protective humoral responses. Nonetheless, the dominant effect in established periodontitis is promotion of neutrophil-rich inflammation and osteoimmune dysregulation leading to connective tissue breakdown and bone loss [100,101]

5.8.2. Experimental Immunomodulatory Strategies to Restore Homeostasis

Given the centrality of the Th17/Treg axis, multiple experimental strategies aim to rebalance adaptive immunity rather than suppress infection. Approaches targeting IL-17 or IL-23 signaling have demonstrated reduced inflammatory infiltration and bone loss in preclinical models, supporting the concept that limiting Th17 effector pathways can attenuate periodontal destruction. Modulation of upstream regulators—such as enhancing A20 (TNFAIP3) activity, a negative regulator of NF-κB signaling—has also been shown to restrain Th17 differentiation and inflammatory amplification [45].

Conversely, strategies that expand or potentiate Tregs represent another promising avenue. Experimental therapies involving IL-35, vitamin D metabolites, mesenchymal stem cells, and regulatory B-cell–mediated IL-10 production have been associated with increased Treg activity and reduced Th17-driven inflammation. These interventions shift the cytokine milieu toward resolution, decrease RANKL expression, and limit osteoclastogenesis. In addition, emerging biomaterial- and nanoparticle-based delivery systems are being designed to locally release immunomodulatory agents within periodontal pockets, enhancing precision and minimizing systemic effects [98,99].

Restoration of periodontal immune homeostasis appears to depend not on global immunosuppression but on recalibrating the Th17/Treg equilibrium. Therapeutic strategies that simultaneously dampen pathogenic Th17 responses and reinforce regulatory pathways hold the greatest potential for controlling inflammation while preserving essential host defense at the oral mucosa [98,99].

5.9. Complement System as a Therapeutic Target

5.9.1. Context and Evolution

The complement system constitutes an interactive network of soluble, cell surface–associated, and intracellular molecules that activate, amplify, and regulate immunity and inflammation. In addition to the classical serum proteins (C1–C9), the system includes approximately 50 proteins, including pattern-recognition molecules, convertases, proteases, receptors, and regulatory components [102]. It comprises three activation pathways—the classical pathway, the lectin pathway, and the alternative pathway—all of which converge at the central component C3 [103].

Activation of the central component C3 generates multiple effector functions, including C3b (opsonization), C3a and C5a (pro-inflammatory anaphylatoxins), and formation of the membrane attack complex (MAC, C5b-9). Beyond its role in antimicrobial defense, complement participates in tissue development, integration of innate and adaptive immunity, clearance of apoptotic cells, and tissue repair [103,104].

However, excessive activation of the complement cascade may contribute to inflammatory pathologies, including periodontitis. Clinical studies have demonstrated increased levels of complement activation fragments in GCF and inflamed gingival tissues.

Periodontal therapy that reduces clinical inflammatory indices is associated with decreased C3 activation in GCF. Conversely, experimental induction of gingivitis increases C3 cleavage, with predominance of the alternative pathway. Reduced expression of the regulatory protein CD59 in diseased gingiva suggests diminished protection against MAC-mediated tissue damage. Genetic and clinical findings further reinforce this association: higher prevalence of the C5 rs17611 polymorphism in periodontitis, C1-INH deficiency linked to aggressive periodontitis with gingival angioedema, downregulation of C3 following periodontal therapy, and identification of C3 among prioritized candidate genes for the disease. Additionally, partial C4 deficiencies are more frequent in patients with periodontitis. These findings support the involvement of complement in periodontal pathogenesis. However, due to their correlational nature, causality was not established, prompting mechanistic investigations in animal models [80].

5.9.2. Complement–TLR Crosstalk

Complement activation is closely linked to Toll-like receptor (TLR) signaling, and both systems often operate simultaneously during microbial infection or tissue injury [104].

Microbial agonists such as lipopolysaccharide (LPS; TLR4), zymosan (TLR2/6), and CpG DNA (TLR9) activate both TLRs and complement, promoting signaling crosstalk in myeloid cells (monocytes, macrophages, neutrophils, dendritic cells). These pathways converge on mitogen-activated protein kinases (ERK1/2 and JNK), which activate key transcription factors (AP-1 and NF-κB), thereby amplifying inflammatory responses [105]. Synergistic activation between C5aR1 and TLR2 in gingival tissues induces significantly higher concentrations of pro-inflammatory cytokines (IL-1β, IL-6, IL-17, TNF) than stimulation of either receptor alone. This synergy appears to be a critical determinant of periodontal inflammation.

Complement hyperactivation in periodontitis and the central role of C3 provide the rationale for developing C3-targeted interventions (e.g., Cp40/AMY-101) as adjuncts to conventional mechanical therapy, with potential to reduce inflammation, modulate dysbiosis, and improve clinical outcomes [106,107].

In preclinical studies in NHPs with ligature-induced or naturally occurring periodontitis, local Cp40 administration reduced clinical attachment loss (CAL), Gingival index, bleeding on probing (BOP), Pro-inflammatory cytokines in GCF (IL-1β, IL-6, IL-8, IL-17A), and alveolar bone loss [108]. Cp40 was clinically developed as AMY-101. Phase 1 trials demonstrated favorable safety and tolerability in healthy volunteers. The FDA subsequently approved a Phase 2a study to evaluate its utility in patients with periodontal inflammation. Intragingival administration on days 0, 7, and 14 resulted in significant reductions in modified gingival index and bleeding on probing, along with decreased levels of MMP-8 and MMP-9, biomarkers of tissue destruction. Beneficial effects persisted for at least 90 days, mirroring the prolonged anti-inflammatory pattern observed in NHPs [109].

Preclinical and clinical data support C3 inhibition via AMY-101 as a promising and well-tolerated host-modulation strategy for the management of periodontal inflammation. The accumulated evidence provides a strong rationale for further evaluation in multicenter Phase 3 clinical trials as adjunctive therapy in inflammatory periodontal diseases [106].

Agnihotri and Gaur (2022) highlighted the therapeutic potential of AMY-101 based on preclinical and clinical evidence, noting its high affinity for C3, prolonged half-life, effective tissue penetration, and sustained complement inhibition after treatment discontinuation. Mechanistically, AMY-101 blocks C3 cleavage, thereby inhibiting complement activation regardless of the initiating pathway and contributing to the restoration of tissue homeostasis. C3-targeted therapy is proposed as a promising adjunct to non-surgical periodontal treatment due to its anti-inflammatory and host-modulatory effects, with additional relevance for patients with systemic inflammatory comorbidities. [110].

5.10. Microbiome and Biofilm Targeted Therapies

5.10.1. Microbial Dysbiosis and Ecological Drivers of Periodontitis

The increasing understanding of periodontal disease as a dysbiosis-driven inflammatory condition—supported by advances in multi-omics analysis, biomarker profiling, and artificial intelligence–based predictive models—has progressively shifted therapeutic strategies toward targeted modulation of the oral microbiome and host–microbe interactions. Within the framework of precision periodontology, these approaches aim not only to control pathogenic biofilms but also to restore ecological balance and regulate inflammatory pathways that sustain periodontal tissue destruction. Periodontitis is currently recognized as a multifactorial chronic inflammatory disease driven by dysbiosis of the subgingival biofilm rather than by the isolated action of a single pathogen. The oral cavity harbors one of the most complex microbial ecosystems in the human body—second only to the gastrointestinal microbiota—comprising bacteria, archaea, fungi, and viruses. Under physiological conditions, these microbial communities maintain a symbiotic relationship with the host and contribute to the ecological stability of the oral environment. However, disruption of this equilibrium by local, systemic, or behavioral factors promotes the emergence of dysbiotic microbial communities capable of sustaining chronic inflammation and progressive destruction of periodontal supporting tissues [111,112].

Current evidence indicates that periodontal disease progression results from the dynamic interaction between polymicrobial biofilms and a dysregulated host immune response. Within this framework, members of the red complex—particularly Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola—play a central role in the ecological reorganization of the periodontal biofilm. Among them, P. gingivalis has been described as a keystone pathogen capable of subverting innate immune responses, facilitating secondary colonization, and creating an inflammatory microenvironment that favors other pathobionts. These microorganisms interact synergistically through virulence factors, proteolytic activity, metabolic cooperation, and immune modulation. In addition, metabolites derived from dysbiotic biofilms—including succinate, propionate, and homoserine—have been associated with epithelial apoptosis, osteoclast activation, and alveolar bone resorption, thereby contributing to the progression of periodontal tissue destruction (Siddiqui et al., 2023; Di Stefano et al., 2023).

5.10.2. Therapeutic Strategies Targeting the Periodontal Microbiome

Conventional periodontal therapy remains primarily based on mechanical disruption of the biofilm, particularly through scaling and root planing, occasionally complemented by local or systemic antimicrobial agents. Although these approaches frequently improve clinical parameters in the short term, their effectiveness may be limited by microbial recolonization and the persistence of dysbiotic ecological niches within the periodontal pocket. Moreover, repeated use of systemic antibiotics raises concerns regarding disruption of the oral and intestinal microbiota as well as the emergence of antimicrobial resistance. For these reasons, contemporary periodontal research has increasingly focused on therapeutic strategies aimed at ecological control of the microbiome and modulation of host inflammatory responses rather than indiscriminate bacterial eradication [112,113]et al., 2023).

Within this evolving paradigm, selective antimicrobial and host-modulatory therapies have emerged as promising adjunctive strategies in periodontal treatment. These approaches include pro-resolving mediators, antimicrobial peptides, inhibitors of the NLRP3 inflammasome, complement C3 inhibitors, anti-adhesion targets directed against P. gingivalis, and exosome-based therapies. Their therapeutic rationale is to interfere with key pathogenic mechanisms—such as microbial colonization, nutrient acquisition, immune evasion, and chronic inflammatory signaling—while preserving beneficial components of the resident microbiota. In particular, inhibition of complement C3 and NLRP3 signaling pathways, together with the use of resolvins and immunoregulatory exosomes, reflects a transition toward immuno-microbial modulation rather than broad antimicrobial suppression [113]

In parallel, increasing attention has been directed toward microbiota-based therapeutic strategies as adjuncts or alternatives to conventional periodontal therapy. Probiotics—particularly strains of Lactobacillus, Bifidobacterium, and Streptococcus salivarius—have demonstrated the capacity to reduce plaque accumulation, gingival inflammation, bleeding on probing, and the relative abundance of periodontal pathogens while exerting local immunomodulatory effects. Prebiotics, synbiotics, and emerging postbiotic approaches aim to promote beneficial microbial communities or reproduce their metabolic functions with improved stability and safety. Furthermore, innovative precision strategies—including bacteriophage therapy, predatory bacteria such as Bdellovibrio bacteriovorus, and oral microbiota transplantation—have been proposed to selectively target dysbiotic microbial communities and restore ecological balance within the periodontal niche. Although many of these approaches remain in early stages of investigation, they highlight the potential for personalized microbiome-based therapies in periodontal care [112,114].

These advances emphasize that periodontitis should be interpreted as the consequence of complex interactions among dysbiotic microbial communities, environmental factors, and host susceptibility. Consequently, future therapeutic strategies are likely to focus on restoring microbial homeostasis and regulating host inflammatory responses rather than solely eliminating pathogenic microorganisms. Such approaches align with the principles of precision periodontology and may ultimately contribute to improved long-term clinical outcomes and more sustainable periodontal health [111,114].

5.11. Advanced Therapeutic Delivery Systems: From Material Passivity to Bio-Intelligence

Periodontal therapeutics has undergone a profound paradigm shift, transitioning from a strict reliance on mechanical debridement to the integration of sophisticated Local Drug Delivery Systems. The inherent limitations of systemic therapies, poor bioavailability within the GCF, sub-therapeutic local concentrations, and systemic side effects have catalyzed the development of biomaterials engineered to overcome the anatomical and physiological challenges of the periodontal pocket [115,116].

5.11.1. Smart and Stimuli-Responsive Hydrogels

Hydrogels serve as the cornerstone of contemporary local delivery innovation due to their high hydrophilicity and ability to mimic the native extracellular matrix. However, current breakthroughs reside in “smart” platforms capable of dynamic responses to the specific microenvironmental cues of the inflammatory site [117].

Thermo-responsive Systems: These polymers exhibit a sol-gel phase transition at physiological temperatures. Formulations based on chitosan and poloxamers allow the material to be injected as a low-viscosity liquid, facilitating adaptation to the irregular morphology of the periodontal defect, before undergoing in situ gelation [115]. This characteristic is vital for resisting the “wash-out” effect caused by GCF turnover and salivary clearance.

pH and Reactive Oxygen Species (ROS) Sensitivity: Periodontal progression is characterized by local acidosis and oxidative stress. Intelligent hydrogels now incorporate pH-sensitive chemical cross-links that degrade selectively in acidic environments, triggered by bacterial metabolism [117,118]. Furthermore, ROS-responsive matrices can act as dual-action systems: scavenging free radicals to mitigate tissue damage while simultaneously releasing anti-inflammatory agents [115].

5.11.2. Nanotechnology: Molecular-Scale Precision

Nanostructured systems have revolutionized the delivery of unstable molecules by enhancing their stability and penetration through the pocket epithelium. These nano-platforms often provide synergistic therapeutic effects beyond simple drug encapsulation [119].

Polymeric Nanoparticles and Liposomes: Poly(lactic-co-glycolic acid) (PLGA) nanoparticles have proven highly effective for the sustained release of antibiotics and host-modulatory drugs, such as statins, which promote alveolar bone homeostasis [116,118]. Liposomes further enable the delivery of fragile peptides and proteins, shielding them from the proteolytic enzymes prevalent in the infected pocket.

Inorganic Nanoparticles (Ag, ZnO, Au): Unlike conventional antibiotics, metallic nanoparticles offer a multi-target mechanism of action that significantly reduces the risk of antimicrobial resistance. For instance, silver nanoparticles (AgNPs) disrupt the respiratory chain of red-complex pathogens, while zinc oxide (ZnO) nanoparticles have been shown to stimulate the proliferation of periodontal ligament fibroblasts [112,119].

5.11.3. Sequential Release and Multimodal Therapies

One of the most disruptive advancements is the design of hierarchical scaffolds that address the distinct phases of periodontal wound healing. Periodontal regeneration requires a coordinated sequence: initial infection control followed by tissue reconstruction [118].

Next-generation scaffolds utilize multi-phasic release architectures:

• Burst Phase: Rapid release of antimicrobial agents to collapse the pathogenic biofilm.
• Intermediate Phase: Controlled release of resolving lipids or anti-inflammatory agents to reprogram macrophage phenotypes from M1 (pro-inflammatory) to M2 (pro-resolving) [77].
• Regenerative Phase: Sustained release of osteogenic factors (e.g., BMP-2, PDGF) or enamel matrix derivatives to recruit and differentiate progenitor cells into cementoblasts and osteoblasts [116,119].

5.11.4. Host Modulation and Microbiome Engineering

Contemporary research has shifted the clinical focus from “pathogen elimination” to “host response management.” Advanced systems are now incorporating disease-modifying antirheumatic drugs (DMARDs) and biological agents that target specific cytokines such as TNF-α and IL-6[77].

Additionally, the localized delivery of prebiotics and probiotics via microencapsulation aims to restore oral symbiosis. By encapsulating strains like Lactobacillus reuteri in bioactive matrices, researchers can ensure bacterial viability against salivary enzymes, allowing these beneficial microbes to outcompete pathogens and re-establish a healthy microbiome [77,112].

5.11.5. Clinical Challenges and Future Outlook

Despite the transformative potential of these systems, translating laboratory findings into chairside clinical practice remains challenging. Barriers include the standardization of degradation rates and the complexity of large-scale manufacturing [120].

Nevertheless, the trajectory toward precision dentistry suggests that “personalized” biomaterials—capable of auto-adjusting drug release based on the patient’s specific inflammatory burden—will define the next decade of periodontal excellence [115].

5.12. Periodontal Therapeutic Algorithms Guided by Advanced Diagnostics

The contemporary evolution of periodontics toward a precision medicine framework necessitates a transition from purely descriptive clinical diagnoses to multidimensional predictive models. Historically, clinical decision-making has relied on static surrogate parameters such as PPD, CAL, and BoP [121]. However, the inherent heterogeneity of periodontal disease and the significant variability in host response limit the capacity of these traditional indicators to forecast intervention outcomes [122] accurately. In this context, the integration of Artificial Intelligence (AI) algorithms and advanced biomarker analysis is redefining therapeutic workflows.

5.12.1. The Paradigm Shift: From Subjectivity to Precision Dentistry

Traditional diagnosis, despite being standardized by the 2017 World Workshop classification, remains heavily dependent on clinician interpretation and historical disease patterns. The introduction of Machine Learning (ML) systems, specifically Random Forest models, now enables the simultaneous processing of demographic variables, radiographic bone loss levels, and systemic risk profiles to suggest optimal therapeutic approaches (surgical vs. non-surgical) with unprecedented accuracy [122]. These algorithms do not merely analyze status; they estimate the probability of pocket closure and long-term tooth loss risk, facilitating a site-specific stratification that guides the periodontist beyond empirical observation [77,122].

5.12.2. Biomarker-Based Algorithms and Host Response Profiling

A critical component of advanced algorithms is the incorporation of the patient’s molecular signature. The analysis of biomarkers in the gingival crevicular fluid (GCF), particularly active-matrix metalloproteinase-8 (aMMP-8), has proven to be an early indicator of disease progression, often preceding clinically detectable tissue damage [78].

Modern therapeutic algorithms propose a dynamic workflow:

o

Multimodal Assessment: Integration of clinical data, AI-analyzed radiographic imaging, and inflammatory biomarker levels.

o

Identification of Immune Phenotypes: Differentiation between patients with a balanced inflammatory response and those with hyper-reactive phenotypes who may require host modulation therapies (HMT) from the initial stages of treatment [77,78].

o

Therapeutic Selection: Determining whether a patient is a candidate for conventional Non-Surgical Therapy (NST) or if anatomical and biological complexities dictate immediate surgical or regenerative intervention to minimize attachment loss [122].

5.12.3. Therapeutic Endpoints and Algorithmic Re-Evaluation

The efficacy of these algorithms is measured by their ability to achieve defined “therapeutic endpoints,” characterized by the absence of deep pockets (≥ 5 mm) with persistent bleeding [121]. AI facilitates this process through predictive monitoring during the re-evaluation phase. If a model identifies a low probability of success post-scaling and root planing, the algorithm can trigger an early transition to surgical phases or the inclusion of therapeutic adjuncts, such as controlled-release agents or systemic host modulators [78]

5.12.4. Challenges and the Future of Guided Diagnostics

Despite the transformative potential of AI-driven platforms, their implementation requires continuous validation across diverse populations. The ability of these models to predict responses in “non-responding” sites—historically the most challenging to manage—remains a core area of development [122]. Nonetheless, the convergence of big data analytics and molecular diagnostics allows for a future where the therapeutic algorithm is not a static guideline but a personalized tool that adjusts in real time to the patient’s biological profile, optimizing tooth preservation and systemic health [121,122].