Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Jun 17;52(9):1288–1297. doi: 10.1111/jcpe.14195

TREM‐1 Pathway Biomarkers for Classification of Periodontal Diseases and Monitoring of Treatment Response in Grade B and C Periodontitis

Angelika Silbereisen 1, Ronaldo Lira‐Junior 2, Beral Afacan 3, Ozgen‐Veli Özturk 3, Gülnur Emingil 4, Nagihan Bostanci 1,✉
PMCID: PMC12377945  PMID: 40528650

ABSTRACT

Background

This study investigated the diagnostic potential of salivary triggering receptor expressed on myeloid cells (TREM)‐1, peptidoglycan recognition protein 1 (PGLYRP1) and interleukin (IL)‐1β for periodontitis patients and their ability to predict treatment outcome.

Methods

Systemically healthy, non‐smokers with gingivitis (n = 31), stage III periodontitis (34 grade B: n = 34, grade C: n = 24) and healthy controls (n = 34) were recruited. Periodontitis patients (n = 42) underwent non‐surgical periodontal treatment. Saliva was collected at baseline (T0) and post‐treatment (T1, T3, T6). Biomarkers were measured using immunoassays. Periodontitis patients were categorised into responders (n = 19) and non‐responders (n = 23) based on the number of residual pockets ≥ 5 mm with bleeding on probing at T6.

Results

TREM‐1 was higher in periodontitis than in gingivitis and health, and in periodontitis grade B than in gingivitis (p < 0.05). PGLYRP1 and IL‐1β were higher in periodontitis and gingivitis compared to controls (p < 0.01). All biomarkers discriminated periodontitis from health (AUC ≥ 0.9) with high sensitivity (82.1%–92.8%) and specificity (83.3%–88.9%). TREM‐1 differentiated periodontitis from gingivitis (AUC = 0.72) with high sensitivity (92.8%) but low specificity (58.1%). Baseline biomarkers did not predict treatment outcome (AUC ≤ 0.61), while T1 levels showed moderate potential (AUC ≥ 0.71).

Conclusions

Salivary TREM‐1 pathway biomarkers offer diagnostic value for periodontitis, are modulated by therapy but show limited ability to predict treatment outcome.

Study Registration

The study has been registered in ClinicalTrials.gov (ID: NCT06715176, 4 December 2024).

Keywords: periodontal treatment, periodontitis, PGLYRP1, saliva, TREM‐1

1. Introduction

Triggering receptor expressed on myeloid cells 1 (TREM‐1) can amplify local and systemic immune and inflammatory responses through regulation of downstream cytokines such as interleukin (IL)‐1β and tumour necrosis factor‐α (Bouchon et al. 2000; Nylund et al. 2018; Tammaro et al. 2017). TREM‐1 is a surface receptor on neutrophils and monocytes but can be released into body fluids as soluble TREM‐1 upon proteolytic cleavage by matrix metalloproteinase or Porphyromonas gingivalis gingipains (Bostanci, Thurnheer, et al. 2013; Bouchon et al. 2000; Gibot and Cravoisy 2004; Weiss et al. 2017). Peptidoglycan recognition protein 1 (PGLYRP1), a neutrophil‐derived bactericidal protein, is one of the ligands that can activate the TREM‐1 signalling pathway, which works in synergy with Toll‐like receptor activation (Bouchon et al. 2000; Dziarski and Gupta 2010; Nylund et al. 2018; Read et al. 2015; Tammaro et al. 2017).

Previous studies have shown that TREM‐1 and PGLYRP1 are elevated in biofluids of patients with various systemic infectious and chronic inflammatory conditions such as, amongst others, sepsis and pneumonia (Gibot, Cravoisy, et al. 2004; Gibot, Kolopp‐Sarda, et al. 2004), rheumatoid arthritis (Inanc et al. 2021; Kuai et al. 2009; Luo et al. 2019) and periodontal diseases (Bisson et al. 2012; Bostanci, Oztürk, et al. 2013; Dubar et al. 2020; Karsiyaka Hendek et al. 2020; Raivisto et al. 2020; Silbereisen et al. 2019). Elevated TREM‐1 levels in gingival crevicular fluid (GCF), serum and saliva of aggressive and chronic periodontitis patients were observed compared to systemically healthy controls (Belibasakis et al. 2014; Bostanci, Oztürk, et al. 2013; Dubar et al. 2020). TREM‐1 also positively correlated with certain clinical periodontal parameters (Bisson et al. 2012; Bostanci, Oztürk, et al. 2013) and periodontopathogens ( P. gingivalis , Treponema denticola and Tannerella forsythia ) in subgingival plaque (Belibasakis et al. 2014). Elevated GCF levels of PGLYRP1 were reported in chronic periodontitis patients compared to healthy controls, both in smokers and non‐smokers (Karsiyaka Hendek et al. 2020). In addition, systemically healthy children and adolescents with gingivitis or subclinical periodontitis presented higher salivary levels of TREM‐1, PGLYRP1 and IL‐1β compared to periodontal health (Raivisto et al. 2020; Yucel et al. 2020). Furthermore, a positive correlation between TREM‐1, PGLYRP1 and IL‐1β serum and/or saliva levels was described (Inanc et al. 2021; Rathnayake et al. 2022; Silbereisen et al. 2019, 2023; Yucel et al. 2020).

However, studies investigating TREM‐1, PGLYRP1 and IL‐1β biomarker levels in response to non‐surgical periodontal treatment are still scarce (Dubar et al. 2020; Karsiyaka Hendek et al. 2020). One study reported significantly reduced TREM‐1 levels in GCF in periodontitis patients after scaling and root planning (SRP) (Dubar et al. 2020). For PGLYRP1, SRP significantly reduced GCF PGLYRP1 levels in periodontitis patients (Karsiyaka Hendek et al. 2020). Furthermore, in an experimental human gingivitis model, salivary TREM‐1, PGLYRP1 and IL‐1β levels were down‐regulated in response to biofilm removal, but the results were not significant (Silbereisen et al. 2019). More studies are needed to clarify the impact of non‐surgical periodontal treatment on the salivary levels of these markers, as well as to investigate whether they are related to clinical response. Thus, this study primarily aimed to investigate the diagnostic accuracy of TREM‐1 pathway biomarkers for periodontitis grade B and C patients. Furthermore, we assessed the biomarkers' ability to predict the outcome of non‐surgical periodontal treatment in those patients. We hypothesised that salivary levels of TREM‐1 pathway markers have the potential to aid in the diagnosis of periodontitis and could predict the response to periodontal therapy.

2. Materials and Methods

2.1. Study Population and Design

For this observational study, systemically healthy, non‐smokers with gingivitis (n = 31), stage III, grade B periodontitis (n = 34), stage III, grade C periodontitis (n = 24) and periodontally healthy controls (n = 34) were consecutively recruited at the Department of Periodontology, School of Dentistry, Aydın Adnan Menderes University, Aydın, Turkey, as previously described (Bostanci et al. 2018). The Ethics Committee of the School of Medicine, Ege University, İzmir, Turkey approved the study (Ethics numbers: 16‐12.1/16, 17‐2/3, 17‐2/4). All participants gave their written informed consent. The study has been registered in ClinicalTrials.gov (ID: NCT06715176, 4 December 2024).

2.2. Clinical Examination

Complete medical and dental histories were taken from each participant with an oral examination. Periodontal measurements included probing depth (PD), clinical attachment loss (CAL), bleeding on probing (BOP), gingival index (GI) and plaque index (PI). All parameters were recorded at six points per tooth, except the 3rd molars, by using a manual periodontal probe (Williams, Hu‐Friedy). The percentage of radiographic alveolar bone loss (RBL) at the interproximal sites was calculated on digital panoramic radiographs as the ratio of the distance between bone level and the cementoenamel junction to the root length.

According to the 2017 diagnostic criteria (Tonetti et al. 2018; Tonetti and Sanz 2019), participants were classified into: (1) generalised stage III, grade C (S3GC) periodontitis; (2) generalised stage III, grade B (S3GB) periodontitis; (3) gingivitis; and (4) periodontally healthy individuals. The specific case definition for each group as well as exclusion criteria is presented in the Supporting Information.

2.3. Saliva Sampling and Non‐Surgical Periodontal Treatment

One week after diagnostic examination, saliva samples were collected and clinical periodontal measurements were recorded in all groups (baseline, T0). The gingivitis group and periodontally healthy controls did not receive periodontal intervention, and clinical measurements and saliva sampling were performed only once (at T0).

Unstimulated whole saliva was collected from the participants via a modification of the saliva collection method as previously described (Navazesh 1993). A description of the collection protocol is presented in the Supporting Information. Saliva samples were stored at −80°C until further analysis.

A subgroup of the periodontitis patients (n = 42 out of 58, 22 grade B and 20 grade C) underwent non‐surgical periodontal therapy by an experienced periodontist (BA) including quadrant‐based SRP one week apart using ultrasonic instruments (Mini Piezon, EMS) and periodontal curettes (Gracey curettes, scaler Hu‐Friedy) until the root surfaces were visibly and tactically clean and smooth as previously described (Greenwood et al. 2020; Silbereisen et al. 2025). The remaining 16 periodontitis patients did not participate in follow‐up treatments. No local or systemic adjunctive agents were administered during SRP. All participants received routine oral hygiene instructions immediately after the first session of SRP and were asked to abstain from any anti‐inflammatory drugs, antibiotics or mouthwashes containing chlorhexidine throughout the study period. Saliva sampling and periodontal measurements were repeated at 1 (T1), 3 (T3) and 6 (T6) months following the treatment completion. At every visit, oral hygiene instructions were reinforced, and the sites that did not respond to treatment at T1 underwent additional re‐instrumentation at T3 and T6.

While non‐surgical treatment efficacy is often reported in terms of mean values of PD reduction and CAL gain, these metrics may not fully capture treatment success or periodontal stability. Therefore, we assessed the presence of residual pockets greater than 4 mm, which serves as a critical risk indicator for disease progression (Feres et al. 2020; Lang and Tonetti 2003). We categorised residual PD ≥ 5 mm with BOP at T6, and periodontitis patients (n = 42) with ≤ 4 sites with PD ≥ 5 mm and BOP were considered as responders (n = 19), while those with > 4 sites with PD ≥ 5 mm and BOP were considered as non‐responders (n = 23).

For subsequent assessment of the biomarkers, saliva samples were available for 34 periodontally healthy individuals, 31 gingivitis patients and 58 periodontitis patients (grade B: n = 34, grade C: n = 24) at T0. Thirty‐six samples were available for analysis at all time points for periodontitis patients. Some samples were not available for analysis (i.e., insufficient sample volume or loss at the time of processing).

2.4. TREM‐1, PGLYRP1 and IL‐1β Immunoassays

For this study, 249 saliva samples were included for the analysis of TREM‐1, PGLYRP1 and IL‐1β. Levels of TREM‐1 (diluted 1:2), PGLYRP1 (diluted 1:100) and IL‐1β (diluted 1:10) in saliva supernatant were measured by commercial enzyme‐linked immunosorbent assays (ELISAs) according to the manufacturer's instructions (all DuoSet ELISA from R&D systems) and as described before (Nylund et al. 2018; Silbereisen et al. 2019). TREM‐1 ELISA detects both membrane‐bound and soluble TREM‐1. Absorbances were measured at 450 nm, with a wavelength correction set to 540 nm to subtract background. Concentrations of TREM‐1, PGLYRP1 and IL‐1β were determined by using a four‐parametric logistic standard curve. Detection limits for the assays, detectability and inter‐assay coefficient of variance are presented in the Supporting Information.

2.5. Total Protein Determination

Total protein levels in saliva were measured by the Pierce BCA Protein Assay according to the manufacturer's guidelines (Thermo Fisher Scientific) by using bovine serum albumin (BSA) standards (range: 25–2000 μg/mL).

2.6. Statistical Analysis

Data analyses were performed using Statistical Package for Social Sciences version 27 (IBM Corporation) and GraphPad Prism version 9 (GraphPad Software Inc.). Continuous variables are presented as mean and standard deviation (SD) and categorical variables as frequencies. Group comparisons were performed with Mann–Whitney, Kruskal–Wallis with Dunn–Bonferroni post hoc or Chi‐square tests, whenever appropriate. Group comparisons before and after treatment were performed with Friedman with Dunn–Bonferroni post hoc test. Correlations between biomarkers and other variables were determined by Spearman correlation coefficient. Area under the receiver operator characteristic (AUC‐ROC) curves were calculated to determine the diagnostic and prognostic values of the biomarkers to identify periodontitis and predict treatment response, respectively. Sensitivity and specificity were determined based on the Youden's index. Differences were deemed statistically significant at p < 0.05.

An a priori power analysis using analysis of variance (ANOVA) for repeated measurements was performed using G*Power 3.1.9.7 to determine the minimum sample size for between‐group (healthy, gingivitis, stage III, grade B periodontitis and stage III, grade C periodontitis) and within‐group (grade B and grade C: between T0, T1, T3 and T6) comparisons. This resulted in a total sample size of 32 participants for between‐group analyses (8 in each of the four groups) and 6 participants for within‐group analyses required to achieve 80% power with a significance of 0.05 and a medium effect‐size of 0.5. TREM‐1 was considered the primary outcome variable of this study.

3. Results

3.1. Demographics and Clinical Findings

This study included participants with a healthy periodontium (n = 34), gingivitis (n = 31) and with periodontitis grade B (n = 34) and grade C (n = 24). Clinical characteristics are described in Table 1. Patients with grade B periodontitis were significantly older than both healthy and gingivitis participants (p < 0.05). As expected, clinical periodontal parameters were significantly worse in grade B and grade C periodontitis in comparison to both healthy and gingivitis (p < 0.05). Additionally, they were also worse in gingivitis than in healthy participants (p < 0.05).

TABLE 1.

Clinical characteristics of participants with a healthy periodontium, gingivitis and periodontitis grade B and grade C.

Healthy (n = 34) Gingivitis (n = 31) Periodontitis grade B (n = 34) Periodontitis grade C (n = 24) p
Age (years) 34.17 (±6.58) 33.16 (±5.85) 42.20 (±5.15) a , b 37.50 (±5.72) < 0.001
Sex (male/female) 16/18 13/18 18/16 11/13 0.846
Plaque index 1.81 (±0.34) 2.85 (±0.54) a 3.54 (±0.56) a , b 3.51 (±0.68) a , b < 0.001
Gingival index 0.03 (±0.01) 1.52 (±0.29) a 2.34 (±0.26) a , b 2.24 (±0.35) a , b < 0.001
BOP (%) 1.57 (±0.92) 68.93 (±12.67) a 85.49 (±11.12) a , b 87.61 (±9.44) a , b < 0.001
PD (mm) 1.69 (±0.16) 2.27 (±0.24) a 3.95 (±0.73) a , b 4.22 (±0.85) a , b < 0.001
CAL (mm) 1.70 (±0.16) 2.28 (±0.24) a 5.17 (±1.15) a , b 5.08 (±1.20) a , b < 0.001
Open in a new tab

Note: Data are presented as mean (±standard deviation) or frequency. p‐value was evaluated by Kruskal–Wallis with Dunn–Bonferroni test or Chi‐square test.

Abbreviations: BOP, bleeding on probing; CAL, clinical attachment loss; PD, probing depth.

a

p < 0.05 in comparison to healthy group.

b

p < 0.05 in comparison to gingivitis group.

3.2. Salivary Levels of TREM‐1, PGLYRP1 and IL‐1β in Periodontal Diseases

We evaluated the levels of TREM‐1, PGLYRP1 and IL‐1β in the saliva of participants with different periodontal diagnoses. The levels of TREM‐1, PGLYRP1 and IL‐1β were significantly elevated in patients with periodontitis grade B and C in comparison with healthy participants (p < 0.01; Figure 1A–C). In addition, levels of TREM‐1 were significantly higher in both periodontitis grade B and C than in gingivitis (p < 0.01; Figure 1A). PGLYRP1 and IL‐1β levels were higher in gingivitis than in healthy participants (p < 0.01; Figure 1B,C). IL‐1β levels were elevated in periodontitis grade C in comparison with gingivitis (p < 0.05; Figure 1C).

FIGURE 1.

FIGURE 1

Open in a new tab

Salivary levels of TREM‐1, PGLYRP1 and IL‐1β in patients with different periodontal diagnoses. Levels of (A) TREM‐1, (B) PGLYRP1 and (C) IL‐1β in healthy (n = 34), gingivitis (n = 31) and periodontitis grade B (n = 34) and grade C (n = 24). (D) Spearman correlation coefficient between biomarkers and clinical variables. Kruskal–Wallis test with Dunn–Bonferroni post hoc test. *p < 0.05; **p < 0.01. Diagnostic accuracy of TREM‐1, PGLYRP1 and IL‐1β for discriminating between periodontitis and (E) healthy or (F) gingivitis. Diagnostic ability was measured by area under the receiver operator characteristic (AUC‐ROC), 95% confidence interval (CI), sensitivity (SN) and specificity (SP). Cut‐off for each biomarker is presented in the figure.

We also assessed the correlation between the salivary markers, age and clinical periodontal parameters. The levels of the three markers in saliva correlated positively with age and with all periodontal parameters (p < 0.05; Figure 1D). TREM‐1 also correlated positively with PGLYRP1 (r = 0.76, p < 0.001) and IL‐1β (r = 0.79, p < 0.001). PGLYRP1 also correlated positively with IL‐1β (r = 0.72, p < 0.001).

In addition, we investigated the diagnostic ability of salivary TREM‐1, PGLYRP1 and IL‐1β in participants with different periodontal diagnoses. All biomarkers discriminated periodontitis from periodontal health (AUC ≥ 0.9, p < 0.001; Figure 1E) with high sensitivity (82.1%–92.8%) and specificity (83.3%–88.9%). Furthermore, TREM‐1 and IL‐1β could also discriminate between periodontitis and gingivitis (AUC ≥ 0.7, p < 0.01; Figure 1F), but with lower sensitivity for IL‐1β (TREM‐1: 92.8%; IL‐1β: 44.6%) and lower specificity for TREM‐1 (TREM‐1: 58.1%; IL‐1β: 87.1%).

3.3. Effect of Periodontal Treatment on the Levels of TREM‐1, PGLYRP1 and IL‐1β

A subgroup of patients with periodontitis (n = 42) was treated with SRP and assessed at T1, T3, and T6 after intervention. Overall, 22 patients with grade B (12 males and 10 females; average 42.6 [±5.1] years) and 20 with grade C periodontitis (9 males and 11 females; average 36.2 [±5.3] years) completed the follow‐up and were included in the final analysis. As shown in Table 2, all periodontal parameters improved significantly at T1 compared to baseline in patients with both grade B and grade C periodontitis (p < 0.05). Furthermore, BOP declined significantly at T3 and T6 in comparison with T1 post‐therapy in grade B periodontitis, whereas it did so only at T6 in grade C periodontitis.

TABLE 2.

Effect of periodontal treatment on clinical periodontal parameters in patients having periodontitis grade B and grade C.

Grade B (n = 22) p Grade C (n = 20) p
Baseline 1 month 3 months 6 months Baseline 1 month 3 months 6 months
Age (years) 42.6 (±5.1) — — — — 36.2 (±5.3) — — — —
Sex (male/female) 12/10 — — — — 9/11 — — — —
Plaque index 3.63 (±0.53) 2.65 (±0.31) a 2.45 (±0.42) a 2.45 (±0.23) a < 0.001 3.37 (±0.56) 2.44 (±0.41) a 2.39 (±0.41) a 2.37 (±0.40) a < 0.001
Gingival index 2.41 (±0.23) 0.65 (±0.33) a 0.50 (±0.27) a 0.47 (±0.27) a < 0.001 2.23 (±0.36) 0.86 (±0.20) a 0.73 (±0.23) a 0.78 (±0.24) a < 0.001
BOP (%) 86.22 (±10.63) 39.90 (±7.93) a 28.93 (±9.04) a , b 27.17 (±9.17) a , b < 0.001 86.54 (±9.79) 38.50 (±13.08) a 32.76 (±10.07) a 30.54 (±10.72) a , b < 0.001
PD (mm) 3.91 (±0.51) 2.70 (±0.61) a 2.56 (±0.54) a 2.57 (±0.59) a < 0.001 4.07 (±0.79) 2.96 (±0.57) a 2.69 (±0.48) a 2.73 (±0.46) a < 0.001
CAL (mm) 5.06 (±0.97) 4.53 (±0.89) a 4.27 (±0.75) a 4.53 (±1.09) a < 0.001 4.78 (±0.94) 4.12 (±0.82) a 3.93 (±0.85) a 4.12 (±0.79) a < 0.001
Open in a new tab

Note: Data are presented as mean (±standard deviation). p‐value was calculated by Friedman with Dunn–Bonferroni test.

Abbreviations: BOP, bleeding on probing; CAL, clinical attachment loss; PD, probing depth.

a

p < 0.05 in comparison to baseline.

b

p < 0.05 in comparison to 1 month.

Then, we investigated the effect of periodontal therapy in patients with periodontitis on the levels of TREM‐1, PGLYRP1 and IL‐1β. For that purpose, saliva was available for all time points for 17 grade B and 19 grade C periodontitis patients. The levels of the three markers decreased significantly 1 month after therapy in grade B patients and remained lower than baseline at T3 and T6 post‐therapy (p < 0.05; Figure 2A–C). In grade C, all three markers reduced significantly at T3 and T6 in comparison to baseline (p < 0.05; Figure 2A–C). Additionally, the levels of TREM‐1 further decreased at T3 compared to T1 after therapy (p < 0.05; Figure 2A).

FIGURE 2.

FIGURE 2

Open in a new tab

Effect of periodontal treatment on the salivary levels of TREM‐1, PGLYRP1 and IL‐1β in patients with grade B and grade C periodontitis. Levels of (A) TREM‐1, (B) PGLYRP1 and (C) IL‐1β before periodontal treatment and 1, 3 and 6 months after treatment in patients with periodontitis grade B (saliva: N = 17) and grade C (saliva: N = 19). Data are presented as mean ± SD. Differences within group were assessed using the Friedman and Dunn–Bonferroni post hoc test. Grade B: a p < 0.05 in comparison to baseline. Grade C: b p < 0.05 in comparison to baseline, c p < 0.05 in comparison to 1 month.

We also analysed whether the levels of TREM‐1, PGLYRP1 and IL‐1β were related to treatment response at 6 months. Periodontitis patients were classified as responders (n = 19) and non‐responders (n = 23) depending on the number of residual sites with PD ≥ 5 mm with BOP. For that purpose, saliva was available for all time points for 17 responding and 19 non‐responding patients. Clinically, both response groups showed a significant improvement in all periodontal parameters at T6 (Table 3). When considering only the patients with saliva available for all time points, the results of the clinical response were essentially the same (Table S1). Regarding the biomarkers, the levels of all of them were significantly higher in non‐responders than in responders at T1 (p < 0.05; Figure 3A–C), which persisted for PGLYRP1 and IL‐1β at T3 (p < 0.05; Figure 3B,C). Importantly, baseline data did not differ significantly between the response groups.

TABLE 3.

Effect of periodontal treatment on clinical parameters in periodontitis patients who responded or not to therapy at 6 months.

Responders (n = 19) p Non‐responders (n = 23) p
Bas 1‐m 3‐m 6‐m Bas 1‐m 3‐m 6‐m
Plaque index 3.5 ± 0.6 2.5 ± 0.3 a 2.4 ± 0.4 a 2.3 ± 0.3 a < 0.001 3.5 ± 0.6 2.6 ± 0.4 a 2.4 ± 0.4 a 2.5 ± 0.3 a < 0.001
Gingival index 2.3 ± 0.3 0.6 ± 0.3 a 0.5 ± 0.2 a 0.5 ± 0.2 a < 0.001 2.4 ± 0.3 0.8 ± 0.2 a 0.7 ± 0.3 a 0.7 ± 0.3 a < 0.001
BOP (%) 86.6 ± 10.3 35.7 ± 10.9 a 26.7 ± 9.8 a 24.8 ± 9.9 a , b < 0.001 86.0 ± 10.0 42.2 ± 9.6 a 33.2 ± 8.7 a , b 32.1 ± 9.0 a , b < 0.001
PD (mm) 3.7 ± 0.4 2.4 ± 0.4 a 2.3 ± 0.3 a 2.3 ± 0.4 a < 0.001 4.2 ± 0.7 3.1 ± 0.5 a 2.9 ± 0.5 a 3.0 ± 0.4 a < 0.001
CAL (mm) 4.6 ± 0.7 3.9 ± 0.6 a 3.8 ± 0.5 a 3.9 ± 0.8 a < 0.001 5.1 ± 1.1 4.6 ± 0.9 4.3 ± 0.9 a 4.6 ± 1.0 a < 0.001
PD ≥ 5 mm + BOP (%) 29.6 ± 9.2 3.5 ± 4.0 a 1.5 ± 1.8 a 0.9 ± 0.8 a < 0.001 48.7 ± 14.8 14.1 ± 9.4 a 10.9 ± 7.2 a 10.2 ± 6.4 a < 0.001
Open in a new tab

Note: Data are presented as mean (±standard deviation). p‐value was calculated by Friedman with Dunn–Bonferroni test.

Abbreviations: BOP, bleeding on probing; CAL, clinical attachment loss; PD, probing depth.

a

p < 0.05 in comparison to baseline.

b

p < 0.05 in comparison to 1 month. Non‐responding patients were those with > 4 sites with probing depth ≥ 5 mm and bleeding on probing 6 months after therapy.

FIGURE 3.

FIGURE 3

Open in a new tab

Salivary levels of TREM‐1, PGLYRP1 and IL‐1β in periodontitis patients according to their response to therapy. Levels of (A) TREM‐1, (B) PGLYRP1 and (C) IL‐1β in saliva before and after periodontal treatment in patients who responded (saliva: N = 17) or not respond to therapy (saliva: N = 19) at the end of the follow‐up. Data are presented as mean ± SD. Differences between groups were assessed using the Mann–Whitney test. *p < 0.05 in non‐responders compared to responders. Prognostic ability of TREM‐1, PGLYRP1 and IL‐1β in saliva measured at (D) baseline or (E) 1‐month post‐therapy for discriminating between responders and non‐responders at the end of the follow‐up. Prognostic ability was measured by area under the receiver operator characteristic (AUC‐ROC), 95% confidence interval (CI), sensitivity (SN) and specificity (SP). Cut‐off for each biomarker is presented in the figure. Non‐responders were those with > 4 sites with probing depth ≥ 5 mm and bleeding on probing 6 months after therapy.

Lastly, we investigated the prognostic ability of salivary TREM‐1, PGLYRP1 and IL‐1β in periodontitis patients responding or not to treatment at T6. The levels of the biomarkers at baseline did not have any significant prognostic value to determine treatment response (Figure 3D). However, when establishing the levels at T1, all biomarkers could discriminate between the treatment response patterns at T6 (AUC ≥ 0.71, p < 0.05; Figure 3E) with moderate to high sensitivity (57.9%–89.4%) and specificity (58.8%–82.3%).

4. Discussion

Measuring TREM‐1, PGLYRP1 and IL‐1β in biofluids has shown potential diagnostic ability in several chronic inflammatory diseases, including periodontitis. However, their modulation in response to treatment and their prognostic ability in predicting treatment outcomes in periodontitis patients are still unknown. We therefore aimed to investigate modulations in salivary levels of TREM‐1, its ligand PGLYRP1 and IL‐1β, a downstream molecule of the TREM‐1 signalling pathway, in periodontally healthy individuals, in gingivitis and periodontitis patients and in response to non‐surgical periodontal treatment.

Our findings show significantly higher salivary levels of all three biomarkers in grade B and C periodontitis in comparison with periodontal health. Furthermore, PGLYRP1 and IL‐1β were able to discriminate between gingivitis and periodontal health, while TREM‐1 was different between periodontitis (grade B and C) and gingivitis. Also, all three biomarkers showed high diagnostic ability, including high sensitivity and specificity, in discriminating between periodontitis and periodontal health when assessing AUC‐ROC characteristics. The association of these three markers with periodontal diseases is in accordance with previous results (Bostanci, Oztürk, et al. 2013; Silbereisen et al. 2019, 2023; Teixeira et al. 2020). We have previously shown that gingival inflammation is one of the main determinants of both TREM‐1 and PGLYRP1 levels in saliva (Rathnayake et al. 2022; Silbereisen et al. 2019, 2023), while pathological periodontal pockets influence TREM‐1 levels (Silbereisen et al. 2023). This could be one of the explanations for the higher levels of PGLYRP1, but not TREM‐1, in gingivitis compared to periodontal health, as well as the higher TREM‐1 levels in periodontitis in comparison to gingivitis. However, in this study, all periodontal parameters strongly correlated with all investigated biomarkers without any distinct association as previously seen. Furthermore, TREM‐1, PGLYRP1 and IL‐1β strongly correlate with each other as previously reported (Rathnayake et al. 2022; Silbereisen et al. 2019, 2023).

It might also be worth mentioning that all three biomarkers were higher in older individuals. A previous similar study only demonstrated an association between age and TREM‐1, but not PGLYRP1 (Silbereisen et al. 2023), maybe as a result of the very large sample size in that cohort. In line with that, age correlated more strongly with TREM‐1 than PGLYRP1 in the current study.

To the best of our knowledge, this is the first time TREM‐1, PGLYRP1 and IL‐1β levels in saliva are investigated in response to non‐surgical periodontal treatment. Our findings showed a substantial reduction of salivary TREM‐1, PGLYRP1 and IL‐1β after non‐surgical periodontal treatment, albeit with differential temporal dynamics in grade B and C periodontitis. In grade B patients, biomarker levels were already significantly lower 1 month after treatment, compared to baseline, while in grade C patients a similar effect could only be observed after 3 months. Interestingly, these differential changes in biomarker levels were not reflected in the periodontal parameters, as both groups showed clinical improvements 1 month after treatment. This suggests a differently regulated TREM‐1 signalling pathway in saliva in these patient groups despite a similar clinical healing process. These results are mostly in line with previous findings obtained from GCF (Dubar et al. 2020; Karsiyaka Hendek et al. 2020) and by use of an experimental human gingivitis model (Silbereisen et al. 2019) where TREM‐1, PGLYRP1 and IL‐1β levels were lower after treatment compared to baseline. A similar study investigating salivary matrix metalloproteinase 8, a biomarker for periodontitis, in response to non‐surgical periodontal treatment in grade B and C periodontitis patients also revealed significant reduction after treatment, however with similar biomarker levels in grade B and C periodontitis patients (Görgülü and Doğan 2022).

We further investigated the potential association between the salivary biomarker levels and their ability to predict the response to treatment after 6 months, based on the number of residual sites with PD ≥ 5 mm with BOP. We found that periodontitis patients not responding to treatment at 6 months presented higher levels of all three biomarkers in saliva at 1 month than patients who responded to treatment, despite similar biomarker levels at baseline. On the other side, clinically, patients in both response groups showed a significant improvement in all periodontal parameters at 6 months after therapy. Thus, measuring the markers 1 month after non‐surgical therapy could offer prognostic value to identify the non‐responding patients. It has been pointed out that a sustained pro‐inflammatory environment leads to periodontitis progression and measuring the levels of inflammatory biomarkers in saliva has the potential to predict disease progression (Kinney et al. 2011; Nagarajan et al. 2019; Teles et al. 2024). Monitoring these biomarkers, particularly in non‐responding patients, may guide personalised treatment strategies and the implementation of host‐response modulation to prevent further tissue breakdown and recurrence (Bostanci et al. 2019; Bostanci and Belibasakis 2012; Donos et al. 2020).

Despite promising, our results should be interpreted with caution. Only periodontitis patients were monitored longitudinally and for only 6 months. Thus, longer follow‐ups in larger patient groups are needed to further investigate the biomarkers' ability to predict disease stability and progression. Also, a feasibility study to evaluate the value of incorporating TREM‐1, PGLYRP1 and IL‐1β levels in the evaluation of a patient' need for retreatment is warranted. It is also worth mentioning that, while the size of the main study groups exceeded the sample size estimates, subgroup analyses in responders versus non‐responders were likely underpowered. Future studies specifically designed and powered for this comparison are warranted to validate these observations. Furthermore, the ability of saliva to reflect the overall oral health status of a patient and not providing any site‐specific information (such as GCF), which might have been more relevant for periodontitis patients, needs to be addressed as well. Also, the biomarker results might have looked differently in stimulated compared to the here‐used unstimulated saliva, due to differences in the saliva's origin and composition. Some strengths of our study are however worth mentioning, such as the inclusion of both grade B and C patients, the thorough clinical examination and the exclusion of potential confounding factors. However, this might reduce the generalisability of our findings, as a large proportion of severe periodontitis cases are found in smokers and diabetics (Silbereisen et al. 2023). Therefore, the predictive ability as well as the external validity of these biomarkers in periodontitis patients with different risk factors should be investigated in future studies.

5. Conclusion

The findings of the present study highlight the diagnostic potential of TREM‐1 pathway markers in periodontal diseases. Although they are modulated by therapy, they do not have any significant predictive value for the periodontal treatment response when measured at baseline. However, we found elevated levels of TREM‐1 pathway markers at 1 month in non‐responders compared to responders, even after a clinically successful treatment, suggesting that clinical improvements may not fully reflect the resolution of inflammatory pathways underlying periodontitis.

Author Contributions

A.S. and R.L.‐J. analysed the data. R.L.‐J. and N.B. conceived and designed the experiments. R.L.‐J. performed the experiments. N.B., B.A. and G.E. designed the study. B.A. and O.‐V.Ö. coordinated and performed patient recruitment, sample collection and non‐surgical periodontal treatments. A.S., R.L.‐J. and N.B. wrote and critically revised the manuscript. All authors approved the work and agreed to be accountable for all aspects of the work.

Ethics Statement

The present study was approved by the Ethics Committee of the School of Medicine, Ege University, İzmir, Turkey, which approved the study (Ethics numbers: 16‐12.1/16, 17‐2/3, 17‐2/4).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting Information.

JCPE-52-1288-s001.docx (60.2KB, docx)

Acknowledgements

We would like to sincerely thank all the participants and the operators who performed the clinical examinations.

Silbereisen, A. , Lira‐Junior R., Afacan B., Özturk O.‐V., Emingil G., and Bostanci N.. 2025. “ TREM‐1 Pathway Biomarkers for Classification of Periodontal Diseases and Monitoring of Treatment Response in Grade B and C Periodontitis.” Journal of Clinical Periodontology 52, no. 9: 1288–1297. 10.1111/jcpe.14195.

Funding: This study was partially funded by the author's institutional funds, the Swedish Research Council (Vetenskapsrådet: 2021‐03528) (N.B.), Karolinska Institute's Strategic Funds and the Steering Group KI/Region Stockholm for Dental Research (SOF) (FoUI‐966140 [N.B.], FoUI‐966258 [R.L.‐J.], FoUI‐990869 [A.S.]).

Angelika Silbereisen and Ronaldo Lira‐Junior contributed equally to this work.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. Belibasakis, G. N. , Öztürk V., Emingil G., and Bostanci N.. 2014. “Soluble Triggering Receptor Expressed on Myeloid Cells 1 (sTREM‐1) in Gingival Crevicular Fluid: Association With Clinical and Microbiologic Parameters.” Journal of Periodontology 85, no. 1: 204–210. 10.1902/jop.2013.130144. [DOI] [PubMed] [Google Scholar]
  2. Bisson, C. , Massin F., Lefevre P. A., Thilly N., Miller N., and Gibot S.. 2012. “Increased Gingival Crevicular Fluid Levels of Soluble Triggering Receptor Expressed on Myeloid Cells (sTREM)‐1 in Severe Periodontitis.” Journal of Clinical Periodontology 39, no. 12: 1141–1148. 10.1111/jcpe.12008. [DOI] [PubMed] [Google Scholar]
  3. Bostanci, N. , Abe T., Belibasakis G. N., and Hajishengallis G.. 2019. “TREM‐1 Is Upregulated in Experimental Periodontitis, and Its Blockade Inhibits IL‐17A and RANKL Expression and Suppresses Bone Loss.” Journal of Clinical Medicine 8, no. 10: 1579. 10.3390/jcm8101579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bostanci, N. , and Belibasakis G. N.. 2012. “Doxycycline Inhibits TREM‐1 Induction by Porphyromonas gingivalis .” FEMS Immunology and Medical Microbiology 66, no. 1: 37–44. 10.1111/j.1574-695X.2012.00982.x. [DOI] [PubMed] [Google Scholar]
  5. Bostanci, N. , Oztürk V., Emingil G., and Belibasakis G. N.. 2013. “Elevated Oral and Systemic Levels of Soluble Triggering Receptor Expressed on Myeloid Cells‐1 (sTREM‐1) in Periodontitis.” Journal of Dental Research 92, no. 2: 161–165. 10.1177/0022034512470691. [DOI] [PubMed] [Google Scholar]
  6. Bostanci, N. , Selevsek N., Wolski W., et al. 2018. “Targeted Proteomics Guided by Label‐Free Quantitative Proteome Analysis in Saliva Reveal Transition Signatures From Health to Periodontal Disease.” Molecular and Cellular Proteomics 17, no. 7: 1392–1409. 10.1074/mcp.RA118.000718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bostanci, N. , Thurnheer T., Aduse‐Opoku J., Curtis M. A., Zinkernagel A. S., and Belibasakis G. N.. 2013. “ Porphyromonas gingivalis Regulates TREM‐1 in Human Polymorphonuclear Neutrophils via Its Gingipains.” PLoS One 8, no. 10: e75784. 10.1371/journal.pone.0075784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bouchon, A. , Dietrich J., and Colonna M.. 2000. “Cutting Edge: Inflammatory Responses Can Be Triggered by TREM‐1, a Novel Receptor Expressed on Neutrophils and Monocytes.” Journal of Immunology 164, no. 10: 4991–4995. 10.4049/jimmunol.164.10.4991. [DOI] [PubMed] [Google Scholar]
  9. Donos, N. , Calciolari E., Brusselaers N., Goldoni M., Bostanci N., and Belibasakis G. N.. 2020. “The Adjunctive Use of Host Modulators in Non‐Surgical Periodontal Therapy. A Systematic Review of Randomized, Placebo‐Controlled Clinical Studies.” Journal of Clinical Periodontology 47, no. Suppl 22: 199–238. 10.1111/jcpe.13232. [DOI] [PubMed] [Google Scholar]
  10. Dubar, M. , Frippiat J. P., Remen T., et al. 2020. “Comparison of sTREM‐1 and Associated Periodontal and Bacterial Factors Before/After Periodontal Therapy, and Impact of Psychosocial Factors.” Journal of Clinical Periodontology 47, no. 9: 1064–1078. 10.1111/jcpe.13339. [DOI] [PubMed] [Google Scholar]
  11. Dziarski, R. , and Gupta D.. 2010. “Review: Mammalian Peptidoglycan Recognition Proteins (PGRPs) in Innate Immunity.” Innate Immunity 16, no. 3: 168–174. 10.1177/1753425910366059. [DOI] [PubMed] [Google Scholar]
  12. Feres, M. , Retamal‐Valdes B., Faveri M., et al. 2020. “Proposal of a Clinical Endpoint for Periodontal Trials: The Treat‐To‐Target Approach.” Journal of the International Academy of Periodontology 22, no. 2: 41–53. [PubMed] [Google Scholar]
  13. Gibot, S. , and Cravoisy A.. 2004. “Soluble Form of the Triggering Receptor Expressed on Myeloid Cells‐1 as a Marker of Microbial Infection.” Clinical Medicine & Research 2, no. 3: 181–187. 10.3121/cmr.2.3.181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gibot, S. , Cravoisy A., Levy B., Bene M. C., Faure G., and Bollaert P. E.. 2004. “Soluble Triggering Receptor Expressed on Myeloid Cells and the Diagnosis of Pneumonia.” New England Journal of Medicine 350, no. 5: 451–458. 10.1056/NEJMoa031544. [DOI] [PubMed] [Google Scholar]
  15. Gibot, S. , Kolopp‐Sarda M. N., Béné M. C., et al. 2004. “Plasma Level of a Triggering Receptor Expressed on Myeloid Cells‐1: Its Diagnostic Accuracy in Patients With Suspected Sepsis.” Annals of Internal Medicine 141, no. 1: 9–15. 10.7326/0003-4819-141-1-200407060-00009.b. [DOI] [PubMed] [Google Scholar]
  16. Görgülü, N. G. , and Doğan B.. 2022. “Effect of Non‐Surgical Periodontal Treatment on Salivary and Serum Biomarkers in Stage III Grade B and C Periodontitis.” Journal of Periodontology 93, no. 8: 1191–1205. 10.1002/JPER.21-0536. [DOI] [PubMed] [Google Scholar]
  17. Greenwood, D. , Afacan B., Emingil G., Bostanci N., and Belibasakis G. N.. 2020. “Salivary Microbiome Shifts in Response to Periodontal Treatment Outcome.” Proteomics. Clinical Applications 14, no. 3: e2000011. 10.1002/prca.202000011. [DOI] [PubMed] [Google Scholar]
  18. Inanc, N. , Mumcu G., Can M., et al. 2021. “Elevated Serum TREM‐1 Is Associated With Periodontitis and Disease Activity in Rheumatoid Arthritis.” Scientific Reports 11, no. 1: 2888. 10.1038/s41598-021-82335-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Karsiyaka Hendek, M. , Kisa U., and Olgun E.. 2020. “The Effect of Smoking on Gingival Crevicular Fluid Peptidoglycan Recognition Protein‐1 Level Following Initial Periodontal Therapy in Chronic Periodontitis.” Oral Diseases 26, no. 1: 166–172. 10.1111/odi.13207. [DOI] [PubMed] [Google Scholar]
  20. Kinney, J. S. , Morelli T., Braun T., et al. 2011. “Saliva/Pathogen Biomarker Signatures and Periodontal Disease Progression.” Journal of Dental Research 90, no. 6: 752–758. 10.1177/0022034511399908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kuai, J. , Gregory B., Hill A., et al. 2009. “TREM‐1 Expression Is Increased in the Synovium of Rheumatoid Arthritis Patients and Induces the Expression of Pro‐Inflammatory Cytokines.” Rheumatology (Oxford, England) 48, no. 11: 1352–1358. 10.1093/rheumatology/kep235. [DOI] [PubMed] [Google Scholar]
  22. Lang, N. P. , and Tonetti M. S.. 2003. “Periodontal Risk Assessment (PRA) for Patients in Supportive Periodontal Therapy (SPT).” Oral Health & Preventive Dentistry 1, no. 1: 7–16. [PubMed] [Google Scholar]
  23. Luo, Q. , Li X., Zhang L., et al. 2019. “Serum PGLYRP‐1 Is a Highly Discriminatory Biomarker for the Diagnosis of Rheumatoid Arthritis.” Molecular Medicine Reports 19, no. 1: 589–594. 10.3892/mmr.2018.9632. [DOI] [PubMed] [Google Scholar]
  24. Nagarajan, R. , Miller C. S., Dawson D., and Ebersole J. L.. 2019. “Biologic Modelling of Periodontal Disease Progression.” Journal of Clinical Periodontology 46, no. 2: 160–169. 10.1111/jcpe.13064. [DOI] [PubMed] [Google Scholar]
  25. Navazesh, M. 1993. “Methods for Collecting Saliva.” Annals of the New York Academy of Sciences 694: 72–77. 10.1111/j.1749-6632.1993.tb18343.x. [DOI] [PubMed] [Google Scholar]
  26. Nylund, K. M. , Ruokonen H., Sorsa T., et al. 2018. “Association of the Salivary Triggering Receptor Expressed on Myeloid Cells/Its Ligand Peptidoglycan Recognition Protein 1 Axis With Oral Inflammation in Kidney Disease.” Journal of Periodontology 89, no. 1: 117–129. 10.1902/jop.2017.170218. [DOI] [PubMed] [Google Scholar]
  27. Raivisto, T. , Heikkinen A. M., Silbereisen A., et al. 2020. “Regulation of Salivary Peptidoglycan Recognition Protein 1 in Adolescents.” JDR Clinical & Translational Research 5, no. 4: 332–341. 10.1177/2380084419894287. [DOI] [PubMed] [Google Scholar]
  28. Rathnayake, N. , Gustafsson A., Sorsa T., Norhammar A., Bostanci N., and Committee, PAROKRANK Steering . 2022. “Association of Peptidoglycan Recognition Protein 1 to Post‐Myocardial Infarction and Periodontal Inflammation: A Subgroup Report From the PAROKRANK (Periodontal Disease and the Relation to Myocardial Infarction) Study.” Journal of Periodontology 93, no. 9: 1325–1335. 10.1002/JPER.21-0595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Read, C. B. , Kuijper J. L., Hjorth S. A., et al. 2015. “Cutting Edge: Identification of Neutrophil PGLYRP1 as a Ligand for TREM‐1.” Journal of Immunology 194, no. 4: 1417–1421. 10.4049/jimmunol.1402303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Silbereisen, A. , Bao K., Wolski W., et al. 2025. “Probing the Salivary Proteome for Prognostic Biomarkers in Response to Non‐Surgical Periodontal Therapy.” Journal of Clinical Periodontology 52, no. 1: 56–67. 10.1111/jcpe.13990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Silbereisen, A. , Hallak A. K., Nascimento G. G., et al. 2019. “Regulation of PGLYRP1 and TREM‐1 During Progression and Resolution of Gingival Inflammation.” JDR Clinical & Translational Research 4, no. 4: 352–359. 10.1177/2380084419844937. [DOI] [PubMed] [Google Scholar]
  32. Silbereisen, A. , Lira‐Junior R., Åkerman S., Klinge B., Boström E. A., and Bostanci N.. 2023. “Association of Salivary TREM‐1 and PGLYRP1 Inflammatory Markers With Non‐Communicable Diseases.” Journal of Clinical Periodontology 50, no. 11: 1467–1475. 10.1111/jcpe.13858. [DOI] [PubMed] [Google Scholar]
  33. Tammaro, A. , Derive M., Gibot S., Leemans J. C., Florquin S., and Dessing M. C.. 2017. “TREM‐1 and Its Potential Ligands in Non‐Infectious Diseases: From Biology to Clinical Perspectives.” Pharmacology & Therapeutics 177: 81–95. 10.1016/j.pharmthera.2017.02.043. [DOI] [PubMed] [Google Scholar]
  34. Teixeira, M. K. S. , Lira‐Junior R., Lourenço E. J. V., et al. 2020. “The Modulation of the TREM‐1/PGLYRP1/MMP‐8 Axis in Peri‐Implant Diseases.” Clinical Oral Investigations 24, no. 5: 1837–1844. 10.1007/s00784-019-03047-z. [DOI] [PubMed] [Google Scholar]
  35. Teles, F. R. F. , Chandrasekaran G., Martin L., et al. 2024. “Salivary and Serum Inflammatory Biomarkers During Periodontitis Progression and After Treatment.” Journal of Clinical Periodontology 51, no. 12: 1619–1631. 10.1111/jcpe.14048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tonetti, M. S. , Greenwell H., and Kornman K. S.. 2018. “Staging and Grading of Periodontitis: Framework and Proposal of a New Classification and Case Definition.” Journal of Periodontology 89, no. Suppl 1: S159–S172. 10.1002/JPER.18-0006. [DOI] [PubMed] [Google Scholar]
  37. Tonetti, M. S. , and Sanz M.. 2019. “Implementation of the New Classification of Periodontal Diseases: Decision‐Making Algorithms for Clinical Practice and Education.” Journal of Clinical Periodontology 46, no. 4: 398–405. 10.1111/jcpe.13104. [DOI] [PubMed] [Google Scholar]
  38. Weiss, G. , Lai C., Fife M. E., et al. 2017. “Reversal of TREM‐1 Ectodomain Shedding and Improved Bacterial Clearance by Intranasal Metalloproteinase Inhibitors.” Mucosal Immunology 10, no. 4: 1021–1030. 10.1038/mi.2016.104. [DOI] [PubMed] [Google Scholar]
  39. Yucel, Z. P. K. , Silbereisen A., Emingil G., et al. 2020. “Salivary Biomarkers in the Context of Gingival Inflammation in Children With Cystic Fibrosis.” Journal of Periodontology 91, no. 10: 1339–1347. 10.1002/JPER.19-0415. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1: Supporting Information.

JCPE-52-1288-s001.docx (60.2KB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Journal of Clinical Periodontology are provided here courtesy of Wiley

RESOURCES