Background
Periodontitis is chronic inflammation of the periodontium caused by the host’s immune response to subgingival bacterial biofilm, which can lead to the irreversible destruction of connective tissue and bone. The gingival epithelium provides the first line of defense against invading bacteria, forming barriers between plaque bacteria and gingival tissue [
1].
The integrity of the junctional epithelium (JE) is therefore essential for maintaining a healthy periodontium. Immunologically, the JE plays a role in protection as a physical, chemical, and immunological barrier to protect the underlying gingival connective tissue and bone from exposure to bacteria and their products [
1]. It has a specialized epithelial structure that attaches the gingival soft tissue to the tooth surface consisting of an internal basal lamina and hemidesmosomes [
2]. During the progress into periodontitis, JE detaches from the tooth surface and migrates apically and laterally toward the space being formed through connective tissue destruction [
3,
4]. Therefore, the detachment of JE from the tooth surface is regarded as the hallmark in the progression of periodontitis.
In previous study, we reported that odontogenic ameloblast-associated protein (ODAM) was extruded from the JE following the onset of JE attachment loss and was detected in gingival crevicular fluid (GCF), and proposed that ODAM could be used as a biomarker of periodontitis and peri-implantitis [
5]. ODAM is a secretory calcium-binding phosphoprotein expressed by ameloblasts during the maturation stage of enamel formation, and its expression persists in the reduced enamel organ and JE of gingiva at the erupted tooth [
6]. It was known that ODAM is implicated in the adhesion of epithelial cells to tooth surfaces [
5,
7]. Wazen et al. recently showed that ODAM plays a role in maintaining the integrity of the JE and gingival healing using an ODAM knockout mouse model [
7]. We also identified that the adhesion of the JE to the tooth surface is regulated via fibronectin/laminin-integrin-ODAM-ARHGEF5-RhoA signaling, and ODAM-mediated RhoA signaling resulted in actin filament rearrangement [
5].
The diagnosis of periodontitis is currently performed using radiography and clinical measurements, such as probing depth (PD), clinical attachment level (CAL), bleeding on probing (BOP), suppuration and mobility [
8]. These traditional clinical measurements not only reflect a history of periodontal disease but also help to determine prognosis, however there is an unmet need for an easily accessible test showing disease activity to diagnose periodontitis.
Extensive research has been carried out on GCF and saliva components that might serve as potential diagnostic markers for periodontitis [
9,
10]. The collection of saliva is relatively simple, safe, and non-invasive, so saliva can be used as a point-of-care diagnostic tool for periodontitis [
10]. However, saliva is limited in detecting disease activity at each individual tooth site, and so traditional clinical measurements should be taken to detect the tooth site affected by periodontitis, even if one subject is diagnosed with periodontitis using saliva. In this respect, the diagnosis using GCF can be useful to diagnose the disease at specific sites. GCF contains a large number of proteins and peptides liberated from the underlying tissues [
9,
11,
12], so the analysis of the GCF components can reflect the disease status of individual tooth sites.
This pilot study sought to investigate the ability of ODAM to reflect the severity of periodontitis at a site-specific level. The analysis of ODAM values from single-site GCF may enable the clinician to distinguish between healthy sites and those affected by periodontitis. For this purpose, we performed a cross-section study of ODAM values in GCF as well as of corresponding clinical parameters in periodontitis patients and analyzed whether there was a relationship between clinical diagnostic parameters and the concentration of ODAM in GCF.
Discussion
This is the first study to assess the possibility of GCF ODAM as a site-specific biomarker for periodontitis. ODAM is involved in the adhesion of the JE to the tooth surface and is released into the gingival crevice when the adhesion is broken by progress into periodontitis [
5]. It was analyzed whether there was a relationship between the value of ODAM in GCF and clinical diagnostic parameters at the same sites. As a result, ODAM appears to serve as a novel site-specific biomarker of periodontitis, as demonstrated by the statistically significant association between the value of ODAM in GCF and the parameters showing the degree of periodontal tissue destruction, PD or CAL. An adjusted linear mixed model showed that the ODAM value in GCF can be an indicator of PD or CAL. Our results suggest that ODAM in GCF plays a role as a site-specific biomarker for deep pockets, although additional studies including the change of ODAM values according to treatment are needed to determine the clinical significance of GCF ODAM. Since the loss of epithelial adhesion from the tooth surface is an early event in periodontal tissue destruction, it is worth verifying the potential of ODAM in GCF as a predictive biomarker for sites that may be vulnerable to periodontal bone loss. Based on the potential of ODAM as a predictive marker for the diagnosis of subjects (using saliva) and sites (using GCF) at risk for periodontitis, development of point-of-care diagnostic tools can help to overcome the limitations of current clinical diagnostics.
Periodontitis is developed at a site-specific level [
17]. Therefore, GCF has been used as a tool for the diagnosis of periodontitis at a site-specific level because it reflects the site-specific severity of periodontitis. It contains a large number of proteins and peptides derived from underlying tissues. To date, nearly 100 different components in GCF have been reported as possible biomarkers for the progression of periodontitis [
9,
12]. These include bacteria or bacterial products [
18,
19], inflammatory mediators [
20], host-derived enzymes and their inhibitors [
21,
22] and soft and hard tissue destruction products [
21,
23,
24]. Although there are many candidates, there has been no validation of factors involved in the attachment of the JE to the tooth surface. ODAM liberated from JE as a result of attachment loss can serve as an indicator of initial periodontal breakdown. Only extremely small volumes of fluid are available from a single site, so GCF requires highly sensitive techniques for quantitative analysis. In this study, ODAM was not detected in 18.9% of the samples. It is necessary to develop a highly sensitive and reliable detection tool that can detect ODAM at low concentrations through the development of a new antibody or aptamer.
The function of ODAM might be related to dentogingival attachment [
5‐
7]. In this study, the value of ODAM in GCF was increased in deep periodontal pockets. This means that pathologic JE can also express ODAM after detachment from the tooth surface. However, in our previous study, ODAM was obviously expressed in the normal JE of healthy teeth but was absent in the pathologic pocket epithelium of diseased periodontium [
5]. ODAM expression was reduced in the JE of experimental periodontitis by drugs, dextran sulfate sodium or periodontopathic bacteria (
Porphyromonas gingivalis) compared with the sham group. Moreover, ODAM was not detected in the pocket epithelium of teeth extracted from periodontitis patients [
5]. It is not readily explained that ODAM, which was not expressed or expressed at low levels in the JE of gingival biopsies from periodontitis patients, was detected at relatively high levels in the GCF from sites with deep pockets. Regarding these results, Wazen et al. noted how the level of ODAM in GCF would be maintained even though it is no longer produced by the JE [
7]. Although the precise mechanism behind the expression of ODAM in periodontitis is unknown, one possibility may be that the detached JE continues to produce ODAM to maintain homeostasis for attachment to the tooth surface, and the resulting ODAM is immediately released into the pocket as soon as it is produced. Therefore, the ODAM may not be observed in histologic specimens of periodontitis models. Since the total area occupied by the pathologic pocket epithelium capable of producing ODAM is increased in periodontitis [
4], the value of ODAM can be increase in deep periodontal pockets. It is also possible that the histological examination of the expression of ODAM according to the severity of periodontitis is not sufficiently verified. In our previous study, the examination of ODAM in JE from human gingival tissue was performed on only one specimen of gingiva around teeth extracted due to periodontitis [
5]. Regarding ODAM expression in the JE of periodontitis, Nakayama et al. showed that ODAM gene expression was increased in inflamed gingiva from patients with chronic periodontitis using DNA microarray [
25]. Recently, they also reported that the expression of ODAM was increased not only at the early stage but also at the following stages in the inflammatory JE on gingival biopsy from an experimental periodontitis model induced by
P. gingivalis. They also showed that the localization of ODAM was spread into the gingival epithelium in inflamed gingiva using human gingival tissues [
26]. To solve these discrepancies, histological examination of the expression of ODAM according to the severity of periodontitis should be performed.
The ROC analysis and AUC calculations were used to assess the ability of the ODAM to reflect sites with PD ≥ 5 mm and positive BOP. Pockets with a PD ≥ 5 mm have clinical significance; when compared with PD < 3 mm, PD ≥ 5 mm represented a risk factor for tooth loss [
16]. Land & Tonetti divided the risk of periodontitis recurrence according to the number of pockets with a PD ≥ 5 mm [
27]. In the analysis based on total subjects, the AUC for ODAM was 0.661, and it was interpreted that the ODAM value in GCF at the least has the potential to serve as a site-specific marker of deep pockets. Additional studies are needed to overcome the low sensitivity and AUC. Considering that the concentration of ODAM in the GCF varies greatly among subjects, it is expected that more accurate cut-off points having high sensitivity and specificity will be determined through the analysis of ODAM from more periodontitis patients.
There are two distinct approaches with respect to reporting GCF mediator content and concentration: the first is to sample GCF for a fixed time period and then report the results either as ρg per 30-s sample or by using the concentration as calculated from the assay (ρg/ml per 30-s sample), and the second is to convert the concentration calculated from the assay back into a concentration based on the original GCF volume according to the Periotron data [
28‐
30]. The second approach allows one to know the actual concentration of the mediator in the GCF; however, it can have the potential for error associated with GCF volume determination and calculation [
28]. A recent review of clinical and technical considerations in the analysis of GCF mentioned that recent authors tend to sample for a fixed period of time (usually 30 s) and report according to the first option described above [
28]. In this study, GCF had to be taken from 40 to 56 sites from each patient, and as it requires a lot of time to measure GCF volume, we adopted the method of reporting by a fixed time period. However, to exclude the possibility that the ODAM value is just reflecting GCF volumes in deep pockets, the relationship between the concentration of ODAM and GCF volume was analyzed from an additional 30 GCF samples collected from 6 independent patients. There was no correlation between the concentration of ODAM and GCF volume (
p = 0.750). The correlation was analyzed using a regression model and Stata/SE 11.1.
A limitation of this study is that GCF samples were collected from sites that have the potential to affect one another in the tooth. The design of the experiments was such that GCF samples were collected from 4 sites from each tooth, and as such, the levels of ODAM in the GCF recovered from mesio (or disto)-buccal sites on a specific tooth cannot be considered to be completely independent from the levels of ODAM in the GCF recovered from the mesio (or disto)-lingual sites on the same tooth. This limitation can lead to errors when analyzing the correlation between the ODAM values in GCF and the clinical parameters. Even so, ODAM in GCF was closely associated with clinical parameters and this can indicate the possibility of ODAM being used as a site-specific marker of deep pockets. Analysis using the GCF samples obtained from completely independent sites can eliminate the related error and provide a more precise correlation. This limitation should be corrected in future studies.
The sample size was calculated that total 424 sites was sufficient to prove the correlation between PD and the value of ODAM in GCF at site-specific level, however the study has the limitation that GCF samples are collected from only 8 subjects. Moreover, the mean ODAM concentration for each periodontitis patient varied. In order to overcome the limitation and apply statistical research methods suitable for data structures, the association between ODAM values and clinical parameters was analyzed using adjusted linear mixed models adjusted with subject effect. As a result, the adjusted model showed that the ODAM value in GCF can be an indicator of PD or CAL.