Biological Markers for Pulpal Inflammation: A Systematic Review

1The dental pulp is equipped to express numerous mediators of inflammation, which can combat irritating factors. Its mechanistic response begins with vascular changes mediated by Toll-like receptors (TLR) 4/2-positive cells and includes release of measurable inflammatory mediators such as IL-8, IL-6, IL-1 and others. Under normal physiologic conditions (left in Fig 1), the vasculature consists of central vessels that branch out into a plexus towards the periphery and specifically the pulp horns. An important difference from soft tissue-enclosed portions of the body is that dental hard tissues enclose the pulp creating a low compliance environment. Dental blood vessels are mainly under control by local metabolites and less by sympathetic innervation. The main cellular components of the pulp are peripherally located odontoblasts and stromal fibroblasts. There are also undifferentiated mesenchymal cells found mainly in the paravascular niche and immune cells (Fig 1). In health, neutrophils predominate but dendritic cells and occasional macrophages are also found.

2Inflammation of the dental pulp (pulpitis) has been viewed as a tightly regulated sequence of vascular and cellular events mediated by molecular factors. Pulpitis is typically caused by an opportunistic infection of the pulp space by commensal oral microorganisms. The most common route of entry for the microorganisms is dental caries. Other potential pathways for pulpal microbial infection include trauma, dentinal cracks, exposed dentinal tubules or the main apical foramen. Cells in human dental pulp that express TLR contribute trigger immune responses to microorganisms and their by-products. This group includes odontoblasts, endothelial cells as well as macrophages and dendritic cells. Some of these cells may form mechanical barriers (i.e. odontoblasts), detect and transmit sensations (nerve fibers) or differentiate (i.e. dental pulp stem cells) to limit infection, signal injury and promote repair, respectively.

3Based on the patients’ signs, symptoms, and examination, four clinical pulpal conditions are described: normal, reversibly inflamed, irreversibly inflamed or necrotic. Histology represents the gold standard to determine the inflammatory state of pulp tissue; however, it is generally agreed that histologic and clinical classification of pulpal diagnosis still needs to be improved and refined. Normal and necrotic pulps have straightforward histological presentation. The conundrum lies in differentiating reversible and irreversible pulpitis. Based on histological reports, reversible pulpitis is characterized by the absence of bacteria and by localized coagulation and liquefaction necrosis immediately surrounding the irritant, whereas irreversible pulpitis is characterized by the presence of the bacteria or their by-products in the dental pulp and by preponderance of acute inflammatory cells predominantly neutrophils in the tissue beneath the lesion suggesting chemotactic activity. Lysosomal enzymes discharged by neutrophils result in widespread tissue damage and suppuration. Acute pulpitis (reversible, and irreversible) can be an extremely painful condition and is believed to be one of the main causes for patients to seek emergency dental treatment during or after office hours. The main clinical difference between reversible and irreversible pulpitis is in the pulp’s response to thermal stimulus. Reversible pulpitis presents an exaggerated yet non-lingering response to cold stimulus. Irreversible pulpitis on the other hand is characterized by constant, spontaneous pain with exaggerated and lingering response to cold stimulus. However, forty percent of teeth with irreversible pulpitis can be painless. In reversible pulpitis, the pulp is expected to recover after removal of the causative stimulus. In contrast, if the pulp is irreversibly inflamed, healing is not expected and pulpectomy (i.e., full removal of the dental pulp) is indicated.

4The succession of signaling events resulting from dental pulp stimulation by microorganisms to the release of an array of immune mediators that in turn may cause pulpal or odontogenic pain, pulpitis, or in advanced stages, pulpal necrosis and finally apical periodontitis have been well described in the past. Detailed discussion of these mechanisms is beyond the scope of this article.

5Currently, diagnostic procedures that aim to assess pulpal inflammation involve case history, as well as clinical and radiographic examination. Clinical examination includes different procedures such as inspection, pulp sensitivity to thermal or electric stimuli, and pain on palpation or percussion. These procedures apparently did not change much in the last century. However, the validity of the currently employed clinical tests to determine the actual or histopathological status of the pulp remains controversial. A recently performed literature review summarized the available information on the diagnostic accuracy of signs/symptoms and current tests used to determine the condition of the pulp. These authors concluded that the overall evidence was insufficient to support the accuracy of such test, even if the tests are combined. Hence, the current diagnostic procedures do not reliably identify the inflammatory status of the pulp. This is particularly unfortunate since decision making in this field, for example differentiation between vital pulp therapy and root canal treatment, critically depends on an accurate pulpal diagnosis.

6According to the National Library of Medicines, the medical subject heading term (MeSH term) definition for a biological marker is a measurable and quantifiable biological parameter that serves as an indicator for health- and physiology-related assessments. Molecules expressed in the cascade of tissue inflammation may serve as (diagnostic) biomarkers for the presence of inflammation. Some research suggests that the dental pulp is not an isolated entity in an encased, solid environment but a reactive tissue that extends its biological products into the outside environment. In fact, studies have shown that pulpal events can be reflected through measurable levels of protein markers that correlated with pulpal symptoms in pulpal blood, dentinal fluid, periapical fluid, and gingival crevicular fluid (GCF;).

7In the field of periodontology, biomarkers in oral fluids/saliva or gingival crevicular fluid are used to detect the occurrence and progression of periodontitis. For example, matrix metalloproteinases (MMPs) such as MMP-8 and -9 have been shown to be central biomarkers of soft tissue breakdown in periodontal pockets. Periodontal and pulpal inflammation shares certain features: initially, both exhibit soft-tissue inflammation caused by microbial infection. At a later stage, these pathologic processes culminate in bone resorption (vertical bone-loss or apical periodontitis, respectively). It is therefore possible that both pathoses may express the same biomarkers. In this regard, MMPs were shown to be potential biomarker for both pulpal and periodontal disease. However, the application of molecular diagnostics in pulpal disease is as yet not used for clinical decision-making.

8Previous studies have investigated the molecular regulatory pathways of pulpal inflammation employing explanted cell cultures in vitro. However, the extrapolation of such results to the clinical situation is difficult, perhaps due to the reductionist nature of such experiments. In vivo, the presence of other cellular players (e.g. immune cells), inhibitory proteins (e.g. protease inhibitors) and other molecules that modify the inflammatory response may present a completely different inflammatory response and consequently, a different clinical outcome compared with what may be suggested by in vitro experimental results. Studies reporting clinical samples for the presence of potential biomarkers for pulpal inflammation are still on the rise. The clinical importance of identifying these biomarkers that can be used to diagnose or to stage pulpal inflammation warrants not only additional studies but also a critical or systematic review and analysis of published reports. Therefore, the aim of this paper is to systematically review the currently available information on biomarkers that were identified from pulp tissues diagnosed as normal or inflamed.

9This systematic review was prepared in accordance with PRISMA (S1 Table). Studies were eligible for inclusion to the review that clinically and/ or histologically differentiate between a healthy and a irreversibly inflamed pulp in permanent human teeth, and analyzed interstitial/ dentinal liquor, gingival crevicular fluid, pulpal tissue, dentin fluid or apical blood for the presence of a biological marker. A biological marker is defined as measurable and quantifiable biological molecule that theoretically can be present in those substrates and might serve as an indicator for a healthy or diseased pulp (adapted from MeSH Unique ID: D015415). An electronic search strategy with combined keywords and indexing vocabulary (MeSH terms) was conducted in the Medline database of the US National Library of Medicine employing the OvidSP interface. We used the following search terms and other subject headings: ‘pulpitis’, ‘acute pulpitis’, ‘irreversible pulpitis’, ‘painful pulpitis’, ‘biological markers’, ‘inflammation mediators’, ‘dentinal fluid’, and ‘gingival crevicular fluid’. S2 Table lists the detailed search strategy performed in Medline. The same electronic search strategy was used in Biosis (OvidSP), the Cochrane library (Wiley), Embase (http://www.embase.com) and the Web of Science (Thomson Reuters). The last date entered was February 19, 2015. No language restrictions were imposed and all articles were included from the inception of the respective database (S3 Table). To ensure the completeness of the search, one reviewer (DRK) conducted a thorough search of the bibliographies of all included studies.

10The search and selection process is summarized in Fig 2. A pool of 1733 records was initially identified using the electronic search strategy and other sources. After removal of duplicates, 851 records remained. Two reviewers (DKR and JCG) independently screened the titles and abstracts of the references collected. Communications not related to the topic were discarded (n = 695). Communications deemed appropriate by one of the reviewers were assigned for full text evaluation. One hundred and fifty-six records were identified using this approach and reviewed as full texts. Articles were collected and evaluated independently by both reviewers. Non-English abstracts or manuscripts were translated with the help of translators. Further articles (n = 99) were excluded for one of the following reasons: i) studies not on human teeth, ii) cell culture study only, iii) no potential biomarker was investigated or the study was off topic, iv) no clear distinction between reversible, irreversible or necrotic pulp, v) studies rather on histologic features or presence of cells, bacteria or viruses than on quantification of a biomarker, vi) review articles, editorials, comments, abstract only or case reports (S4 Table). In case of disagreement consensus was achieved through discussion by third party arbitration (OAP). Articles where no exclusion criteria applied were included to the review. There was 94.2% agreement prior to arbitration between both reviewers and finally 57 publications were included to the review. The included articles were written in English (n = 54) or Chinese (n = 3) language.

11The quality of the included studies was assessed using a modification of the Newcastle-Ottawa-Scale (NOS;). The NOS rates the 3 study domains ‘selection’, ‘comparability’ and ‘outcome’. Each positive rating was awarded with a star. The parameters recorded for ‘selection’ were: selection of the cohort (gender and age distribution reported) and condition of the cohort (general health and medication reported). The parameters recorded for ‘comparability’ were: diagnostics of cases and controls (anamnesis, clinical and radiological inspection described in sufficient detail), histological confirmation of the diagnosis performed (yes/no), quality of the controls (control sample from the same patient as the case sample) and the ratio of the group size (cases:controls ≥ 1:2). The parameters recorded for ‘outcome’ were: reported blinding to the case/control status (yes/no) and that the same tests were performed with the cases and control samples (yes/no). Consequently two stars could be awarded for ‘selection’, four stars for ‘comparability’ and two stars for ‘outcome’. A value of ‘0’ represents the lowest study quality and ‘8’ the highest possible quality rating of the modified NOS.

12Quantitative data were collected from all studies included to the review. An electronic protocol for data extraction was defined and piloted on several manuscripts before final completion. Relevant information regarding reference name, publication date, substrate analyzed for the presence of a biological marker, how was the substrate collected, number of specimens in experimental group and control group, was the substrate pooled before analysis (yes/no), name of the biological marker under investigation, what type of molecule is the biological marker, what general function serves the biological marker, what was the molecular expression level of the biological marker, which analyte was evaluated for the presence of the biological marker, which method was used for analysis, and were statistically significant differences between specimens of the irreversible pulpitis group compared to the control group (healthy pulp) reported (yes/no), were collected. The synthesis of the data is presented in a descriptive manner. Moreover, descriptive statistics were applied when deemed appropriate (JMP 10.0.0, SAS Institute, Cary, N.C., USA).

13The studies excluded during full text evaluation (N = 99; Fig 2) are presented in S4 Table. The studies included in the review are listed in Tables 1 and 2. Due to the heterogeneous nature of the studies it was not possible to perform meta-analysis on their outcome. The quality ratings of the included studies according to the modified NOS are presented, along with the full reference, in S5 Table. The average quality score was 3.9 ± 1.1 (mean ± SD). Fig 3 provides an overview on the total ratings for the respective parameters for the study domains selection, comparability and outcome. Weaknesses were noticed for the parameters selection of the cohort, quality of the controls and reported blinding to the case/control status.

14Eighty-eight percent of the studies included (50/57; Table 1) analyzed pulp tissue for the presence of a biomarker either collected via pulpectomy (N = 5), tooth extraction and fracturing (N = 25), or a combination of both (N = 20). Twelve percent (7/57; Table 2) of the studies included analyzed substrates other than pulp tissue: pulpal blood (N = 2), peripheral blood serum (N = 1), GCF (N = 1), dentinal fluid (N = 1), or extracellular pulpal fluid (N = 2). Pulpal blood, GCF and dentinal fluid were collected using absorbable membranes, blood serum via peripheral blood collection and extracellular pulpal fluid by inserting microdialysis membranes into vital pulp tissue. Eighty-two percent of the studies analyzing pulp tissue (41/50) used tissue collected from extracted healthy, non-carious permanent, or wisdom teeth as their control. Fourteen percent (7/50) used healthy pulp tissue collected via pulpectomy because of elected root canal treatment for prosthetic reasons as their control. One study used tissues from extraction and pulpectomy as control, another one did not state precisely how they collected control tissue. Of the 7 studies evaluating substrates other than pulp tissue, two sampled blood, and another one extracellular fluid from healthy teeth that were assessed but subsequently planed for extraction because of prosthetic or orthodontic reasons as control. One study sampled venous (peripheral) blood during pulp inflammation and used a consecutive peripheral blood sample after treatment as control. One further study sampled GCF from healthy contralateral or adjacent teeth as control, and another one collected dentinal fluid from non-symptomatic teeth scheduled for replacement of a filling as control. The substrate in one study was pooled before performing the confirmatory test.

15Analytical methods used for the assessment of pulp tissue included reverse transcription polymerase chain reaction, multiplex assay, microarray, western blot, radioimmunoassay, immunohistochemistry, enzyme-linked immunosorbent assay, zymography, flow cytometry, limulus amoebocyte assay and specific enzyme assays (Table 1). Pulpal inflammation was confirmed by histology in 42% (21/50; Table 1 and S5 Table) of these studies. Substrates other than pulp tissue were analyzed using radioimmunoassay, enzyme-linked immunosorbent assay, specific serum, or enzyme assays (Table 2). Histology was not used to confirm pulpal diagnosis in those studies. Seventy-four percent of the studies evaluating pulp tissue (37/50) analyzed actual protein expression or protein activity, whereas 16% (8/50) analyzed the pulp tissue on the DNA level. Five studies (10%) analyzed the pulp tissue substrates at both levels (Table 1). All studies evaluating other substrates than pulp tissue evaluated protein expression or protein activity (Table 2).

16Pulp tissue was assessed for a total of 89 biological markers. Statistical significant differences between an irreversible inflamed and a healthy pulp could be detected for 64 biological markers (71.9%) by at least one study. Nineteen biological markers showed no statistically significant differences between inflammation and health, whereas 6 biological markers were not evaluated employing statistical tests (Table 1). Substrate other than pulp tissue was evaluated for 16 biological markers. For twelve biological marker (75%) statistical significant differences between irreversible inflammation and health could be detect by at least one study.

17The results point to a response in pulpitis by immunocompetent tissues that ultimately results in the release of mediators, which in turn trigger a series of inflammatory events and an attempt to initiate repair. Collectively, the data presented here demonstrate the involvement of various TLR-induced chemotactic molecules (i.e. IL-8, CXCL-10, MIP family, GRO, MCP family, RANTES, Eotaxin, IP10, and others).

18TLRs have been shown to confer immunocompetence to the dental pulp. They are expressed by both immune and non-immune cells in the pulp including neurons, fibroblasts, endothelial cells, epithelial cells and others, which recognize viral and microbial structures as well as self molecules (such as single stranded RNAs) that may accumulate in non-physiologic amounts or sites during inflammation.

19Under normal conditions, very few immune cells are present in the dental pulp. In the presence of infection (i.e. caries), immune cells are recruited to the pulp even in the absence of direct bacterial contact on the pulp tissue itself. The permeability of dentin to soluble bacterial products allows pulpal response to occur prior to carious pulpal exposure. These soluble bacterial products, along with components of the complement system and products of the lipoxygenase pathway of arachidonic acid metabolism are chemotactic for leukocytes.

20The exponential increase in the number of infiltrating leukocytes brings with it a corresponding increase in lysosomal enzymes that cause tissue damage. Proteases like elastase and MMPs (Tables 1 and 2) cleave elastin and proteoglycans that destroy the pulp tissue resulting in irreversible damage. Furthermore, the accompanying spike in inflammatory mediators like PGE2, cAMP, COX-2, CGRP, neurokinins and others stimulate vasodilation and microvascular permeability by binding into their respective receptors (i.e. EP2/3 receptor for PGE2) and induce cytoskeletal rearrangement or contraction of vascular smooth muscle.

21Equally as important is the action of neuropeptides (e.g. substance P, calcitonin-gene related peptide) (Table 1). These neuropeptides typically reside in endings of afferent nerve close to blood vessels but also associated with macrophages and odontoblasts. As a response to stimuli, afferent nerve sprouting has been demonstrated, and with it an increase in neuropeptide concentration, which can cause spontaneous pain, allodynia or hyperalgesia in teeth with pulpitis.

22Simultaneous to the destructive effects of leukocytic infiltration is the capability of these cells to induce repair through the release of VEGF, TGF-B, GM-CSF and others (Tables 1 and 2) that induce alterations of the local extracellular matrix, promote induction of endothelial cells to migrate or proliferate, and inhibition of vascular growth with formation of differentiated capillaries. The increased expression in inflamed pulp of toll-mediated human beta-defensins (hBD) that play an important role in the innate host defense against bacterial invasion, contribute to promotion of adaptive immune responses, and show chemotactic activities further underscore the dynamic range of response of the dental pulp during inflammation. In addition, it can also be appreciated that during pulpal inflammation, the anti-inflammatory effects of various mediators such as tissue inhibitors of matrix proteinases (TIMPs), siRNA and others also come into play.

23As a direct result of the release of inflammatory biomarkers, pulpal responses include classical signs of inflammation specifically a vascular response, along with changes in mediator profiles and cellular constituents. The transition from reversible to irreversible pulpitis has been broadly characterized by a migration of dendritic cells towards odontoblasts and accumulation of immune cells. However, a more detailed analysis such in the majority of studies included in this paper evaluating biomarkers of pulpal inflammation demonstrates (statistical) significant differences between a clinically diagnosed healthy or irreversibly inflamed pulp at the molecular level.

24Moreover, the analytes were obtained via different approaches both from the pulp directly as well as indirectly from tissue fluid. Fig 4 illustrates the potential sampling sites for molecular pulpal diagnostics. While having the benefit to show a direct picture of intrapulpal conditions using pulpal blood or whole pulp tissue requires access to the pulp space and is therefore not applicable as a chairside screening tool. Conversely, indirect methods such as dentin fluid collection or assessment of mediators in GCF can be performed clinical in a less invasive way. Dentin fluid is the extracellular fluid that is contained within dentinal tubules; its composition includes inflammatory mediators and vasoactive compounds associated with inflammation. While initial evidence suggested that these mediators can be assayed problems exist with protein yield and the need to remove the existing restoration or in other cases to prepare an initial cavity deep in dentin.

25GCF was used to sample mediators in one study included here (Table 2). This fluid is an exudate that from the gingival crevice and it contains several host factors including antibodies, bacterial antigens, proteins, and cytokines. GCF analyses may be promising due to the ease of application. Moreover, it may be possible to assess the dynamics of apical periodontitis using GCF. However, the major drawback in the evaluation of host mediators in GCF is that tissue inflammation, independent of its cause, is a non-specific process of innate immunity and this makes it difficult to distinguish on a molecular level between a marginal and apical periodontal inflammation. When sampling from the GCF for pulpal diagnostics, this potential drawback could be overcome by (i) creating healthy periodontal conditions, (ii) averaging out several sites on one or multiple teeth, (iii) combining clinical and radiographic observations, or (iv) defining a specific pattern of metabolites relevant to the pulp and not the periodontium, or possibly by other as yet unidentified means. Furthermore, the detection of mediators of pulpitis in GCF is impacted by the need for these compounds to reach the periodontal ligament and ultimately the gingival crevice in sufficient concentrations. Indeed, the notion that mediators will diffuse from the pulp via dentinal tubules or accessory canals to the periodontal ligament has been called into question. Periapical fluid samples, while requiring a direct access to the apical site, are of interest to determine the level of systemic inflammation.

26Discovering an improved method to determine the present inflammatory condition of the pulp could be of great value: on the one hand, pulp necrosis is one of the most frequent complications after coronal restoration of assumed non-inflamed (vital) teeth, on the other performing a full pulpectomy on teeth that could have been kept vital (at least in part) suggests that overtreatment may occur in many cases. Endodontic diagnosis should therefore focus on either the extent of the microbial infection or the inflammatory reaction of the host tissue; however, current methods do neither.

27Keeping a pulp vital offers distinct advantages compared to root canal treatment: the protective immune capacity of the pulp remains preserved and the remaining tooth structure gets not unnecessarily weakened by access cavity preparation and root canal enlargement. Unfortunately, the only available long-term outcome studies on direct pulp capping procedures (i.e. direct pulpal interventions), which attempt to maintain pulpal vitality, show unsatisfactory success rates as low as 20% after ten years. The development of biocompatible materials facilitates a wound closure free of inflammation after pulpal capping procedures or partial pulpotomy. However, the likelihood of a pulp to survive such procedures remains questionable using current schemes for assessment of pulpal inflammation.

28One limitation of this systematic review is that merely 2 out of 57 studies were specifically designed to investigate potential biomarkers in the context of pulpal diagnostics. Most of the studies analyzed here merely target the presence of molecules and their function in pulpal inflammation. Nevertheless, based on the current state of knowledge this review provides an overview on molecules that are present and measurable during pulpal inflammation and therefore potentially can serve as a biomarker for pulpal inflammation. This may provide impulses for further research. This research needs to explore patient (age, gender, systemic condition) and infection related factors (varying composition of the microbiological infection). Clinical investigations should be conducted that are specifically designed to confirm the results collected from the research collected here. More specifically mediator profiles should be assessed in defined clinical scenarios. In addition, the assays methodology should be tested for their applicability with the possible substrates. The ultimate goal should then be to develop an inexpensive chairside test for non-invasive molecular pulp diagnostics. In fact, such a chair-side assay, based on the immunochromatographic detection of MMP-8 specific antibodies, is already commercial available to diagnose periodontal inflammation. For endodontic procedures of the future, such as partial pulpotomies and pulp regeneration, a comparable test will be of significant value.

29Indeed, various biomarkers that are produced by cellular components of the dental pulp can provide a snapshot of the biological mechanisms that propel this immunocompetent tissue towards healing or necrosis. The imbalance between tissue destructive molecules like proteases and tissue inductive molecules like VEGF may serve as a diagnostic or prognostic tool for endodontic intervention. The challenge remains on developing a method to make these biomarkers readily measureable in a clinical setting.

30In the included studies, irreversible pulpitis was associated with different expression of various biomarkers compared to non-inflamed controls. These biomarkers were significantly expressed not only in pulp tissue, but also in gingival crevicular fluid that can be collected non-invasively and in dentin fluid that can be analyzed without extirpating the pulpal tissue. This may be used to accurately differentiate diseased from healthy pulp tissue. The main current challenges in the clinical application of biomarkers lie in the identification of biomarkers or biomarker subsets that reliably correlate with pulpal inflammation, the improvement of sample collection (substrate and protein yields), and their analysis (interference of the biomarkers with inflammation of other than pulpal origin). If these hurdles can be overcome, a more accurate pulpal diagnosis and more predictable vital pulp treatment regime may create better clinical outcomes.