1 INTRODUCTION

The high prevalence of severe periodontal disease globally1 and its consequences with regard to reduced esthetics,2-4 function2-4 up to tooth loss (TL),2-4 and the resulting decrease in oral health‒related quality of life5 underline the significance of understanding and examining the multifactorial nature of periodontitis2, 3, 6, 7 on a clinical level in addition to basic research.

If patients have successfully undergone active periodontal therapy (APT), application of a systematic periodontal treatment concept encompasses allocation to supportive periodontal care (SPC).8, 9 After completion of APT, it can be assumed that TL during SPC, depending on its duration, is a rare event.6-13 Today, the general relevance of SPC to facilitate long-term tooth retention is widely acknowlegded.6-15 However, only a minor portion of these studies consider that SPC is a fundamentally periodontal treatment method, which primarily aims to prevent TL for periodontal reasons (TLP).7, 14, 15 It is therefore important that studies on this topic cover long SPC periods and consider TLP so as to better assess the success of SPC. This is not easy, as there is no standardized definition of periodontally hopeless teeth. On the contrary, the limits of long-term tooth retention seem to be increasingly shifting to longer survival.16 A large number of retrospective cohort studies in both, a university6-8, 10-14, 17-21 as well as in a private practice setting2, 15, 22, 23 already exist, further long-term data are valuable, as they help to confirm and expand existing evidence and to incorporate therapy results of different concepts at different centers.

A distinction is made mainly at the level of data analysis between patient- and tooth-related data. However, prognostically, these data must always be considered together within a patient. The aim of this retrospective cohort study was to identify tooth-related factors (data on patient-related factors published previously7) for overall TL (OTL) and TLP in a homogeneously treated cohort over a period of 10 years.

2 MATERIALS AND METHODS

Patient-related data from this cohort have recently been reported.7 Selected data of the first 50 patients from this study were considered in a multicenter project.8

2.1 Patients After patients were identified by electronic and manual database searches by means of dental codes. They were invited consecutively, in the order of their treatment at that time, to a follow-up examination that took place 120 ± 12 months after APT completion until a number of about 100 patients was included.3, 13, 22 Only those patients who had undergone anti-infective therapy after 2005 in the Department of Periodontology of the Johann Wolfgang Goethe-University Frankfurt/Main were considered for possible inclusion in this study. Further inclusion criteria were as follows: Complete periodontal status [periodontal probing depth (PD), clinical attachment level (CAL), bleeding on probing (BOP) at six sites/tooth; furcation involvement (FI)24 at all furcation sites of multi-rooted teeth; tooth mobility25 of all teeth before start of therapy (baseline, T0), and after completion of APT (non-surgical/step 2 and, if required, surgical/step 3 therapy9), and start of SPC (T1) Modified full-mouth disinfection concept (FMD)26 was applied Age ≥18 years at the time of re-examination (T2) Panoramic radiograph or complete set of periapical radiographs from baseline T1−T2 = 120 ± 12 months

This study was approved by the Institutional Review Board for Human Studies of the Medical Faculty of the Johann Wolfgang Goethe-University (approval no. 61/15) and was conducted in accordance with the 1975 Declaration of Helsinki, as revised in 2013. All patients gave written consent for participation in this study. The study was registered with the United States National Library of Medicine clinical trials database (ClinicalTrials.gov; ID: NCT03048045).

2.2 Applied treatment concept The treatment concept adopted during the course of this study has already been described in detail.7, 26 At the beginning, all subjects received oral hygiene instructions and supragingival instrumentation. Afterward, subgingival instrumentation according to a modification26 of the FMD concept was performed.27 FMD was combined with adjunctive systemic antibiotics if Aggregatibacter actinomycetemcomitans was detected by different commercially available sets for taking subgingival plaque samples.* ,† ,‡ If required (e.g., remaining PD ≥6 mm9), periodontal surgery was recommended. After completion of APT, patients were allocated to SPC according to the Periodontal Risk Assessment.28 As a result, the risk-adapted SPC interval was determined prospectively in each individual SPC session.3 This concept anticipates in large parts the actual clinical practice guideline for treatment of stages I, II, and III periodontitis.9 SPC was encompassed consistently:7

Gingival bleeding index (GBI)29 and plaque control record30 at six sites/tooth

Re-instruction and re-motivation for effective individual biofilm control

Professional mechanical plaque removal

Application of fluoride gel* 31

Twice per year, a dental examination and a complete periodontal status assessment including PD, BOP, FI, and tooth mobility (some teeth have been splinted since T0) were recorded. Once per year, CAL was assessed. At sites with PD = 4 mm + BOP or PD ≥5 mm,32 subgingival instrumentation was performed and 1% chlorhexidine digluconate gel† was instilled.

All treatments were performed in a university setting by dentists in collaboration with dental nurses or hygienists as well as by dental students under supervision of periodontists and postgraduate periodontics students. If a patient exhibited >5 teeth with PD ≥5 mm 2 years after re-evaluating APT, recurrence therapy was recommended, considering individual factors (e.g., patient age, systemic diseases).

2.3 Variables evaluated at 10-year re-examination (T2)

Four experienced periodontists (KN, TR, PE, HP) were involved in patient re-examinations between July 2015 and April 2019. Interindividual calibration for PD and CAL by repeated measurements has been described in detail previously.7

Tooth-related factors included the following:

PD and CAL to the nearest 1.0 mm with a manual, millimeter-scaled rigid periodontal probe‡ at six sites/tooth

BOP reported 30 seconds after probing

FI at all multi-rooted teeth with furcation probe§ 24

Tooth mobility at all teeth25

Dental status (assessment of teeth lost during SPC, tooth type [anteriors, premolars, molars], and abutment status [no abutment tooth, fixed, or removable partial denture])

Patients who lost teeth during SPC were asked about the reason for this, if teeth were removed outside of the center.

The following patient-related factors were considered: 1) Self-reported smoking status (non-smoker [never smoked in their life], former smoker [stopped smoking ≥5 years ago], active smoker [including patients who stopped smoking <5 years ago]);28 2) Medical history; and 3) GBI29 and plaque control record.30

2.4 Radiographic examination

A panoramic radiograph or a complete set of periapical radiographs was available for each patient at T0 by one non-calibrated examiner (LP). Each tooth was assigned the highest mesial or distal bone loss (BL) according to one of five categories (≤20%, 21% to 40%, 41% to 60%, 61% to 80%, >80%) using a Schei Ruler.33

2.5 Variables evaluated using patients’ charts

This study evaluated the following: 1) Medical history (diabetes status [including HbA1c]; smoking status [including number of cigarettes/day]); 2) Initial diagnosis of periodontal diseases (1999 classification),34 reclassified according to the 2018 classification using T0 periodontal charts (staging: interproximal CAL, teeth missing due to periodontal reasons and complexity; grading: BL age index, smoking, diabetes);35 3) Periodontal charts between T1 and T2 (PD, CAL, BOP, FI, tooth mobility, for calculation of periodontal inflamed surface area [PISA] and periodontal epithelial surface area [PESA]36); 4) GBI29 and plaque control record;30 5) Adherence (adherent/non-adherent) by comparing SPC interval recommendations with intervals actually documented in the patient's file (if patients once exceeded the recommended SPC interval by > 100%, they were considered to be non-adherent3); 6) SPC period and number of SPCs; and 7)Reasons for TL if tooth/teeth were removed in the authors’ center (TLP: if a combination of progressive CAL loss, FI II/III,24 and/or tooth mobility II/III25 was found); since the documentation of extraction decisions over the past 10 years has not been uniform, the last clinical and radiological findings before the respective extraction were used—if reasons were not explicitly documented—to assess whether there were either periodontal or other reasons for TL.7

2.6 Statistical analysis

Data of all subjects at T0, T1, and T2 were entered into a data matrix.* The patient was defined as the statistical unit and OTL/TLP during SPC as the main target variable. Third molars were excluded from data analysis.

Patient-specific characteristics were described using absolute (mean ± SD) and/or relative frequencies. Tooth-specific data were described separately at T0, T1, and T2 using absolute and relative frequencies. Univariate correlations of patient-related variables were performed for metrically scaled data using Pearson correlation coefficient and for nominally scaled data using the Chi-square test. Tooth-related data were compared by repeated-measures analysis of variance.

Two logistic multilevel regression models were calculated using “OTL” or “TLP” as dependent variables (“0” = tooth not lost; “1” = tooth lost) to identify tooth-related factors possibly affecting TL. Therefore, the level “teeth” (T1) was subordinated to the level “patient.” As an indicator of how well the model fits the data, “-2 log likelihood” (-2LL) was calculated. Factors were included by significant binary logistic regression analysis. Collinearity was tested for all independent variables/factors by variance inflation factor showing VIF <2.1.37 As tooth-related variables, abutment status, FI (most severe score per tooth), tooth mobility, BL, mean PD, and CAL of six sites/tooth were considered (all T1). As patient-related factors, mean BOP, GBI during SPC (for OTL and TLP), the number of SPCs, grading, and smoking at T1 (for TLP) were included into the model. PISA, PESA, and tooth type were not considered in the regression model due to collinearities (VIF >10).37

A significance level of 0.05 was assumed. The statistical evaluations were performed with appropriate software.†

3 RESULTS 3.1 Patients

Among 153 consecutively screened patient files, 45 patients were excluded due to violation of inclusion criteria. Twelve did not receive FMD, but instead received quadrant-wise subgingival instrumentation and were therefore excluded by deviation from the treatment concept. Of the remaining 108 patients, four denied participating in the study, six were no longer available at their known addresses, and one patient was deceased (Fig. 1).

Patient flow diagram

Thus 97 patients (n = 51 female; 53%) with an average age of 55.2 ± 10.9 years at T1 were included. SPC lasted 10.2 ± 0.5 years on average. Fourteen patients (14%; T1) were smokers and five (5%, T1) suffered from diabetes. Eleven patients received systemic antibiotics adjunctively to subgingival instrumentation (APT) (11%), 13 underwent recurrence therapy (13%), and 55 regularly participated in SPC (57%). Patients with TLP attended on average six appointments more than patients without TLP. Further patient-related data are depicted in Table 1.

TABLE 1. Patient characteristics Variables Total All patients without OTL All patients with OTL P All patients without TLP All patients with TLP P Patients, n 97 50 47 0.452 74 23 0.452 Sex, n (female/male) 51/46 28/22 23/24 0.486 39/35 12/11 0.965 Age, years T0 54.07 ± 10.90 53.19 ± 10.73 55.02 ± 11.12 0.412 54.11 ± 10.83 53.97 ± 11.37 0.957 T1 55.16 ± 10.88 54.15 ± 10.65 56.22 ± 11.12 0.351 55.11 ± 10.75 55.32 ± 11.51 0.935 T2 65.33 ± 11.0 64.33 ± 10.68 66.39 ± 11.15 0.355 65.27 ± 10.81 65.52 ± 11.43 0.924 Smoking, n (T1) Active smoker 14 6 8 0.654 10 4 0.030 Former or non-smoker 83 44 39 64 19 Diabetes, n 5 2 3 0.668 3 2 0.768 APT Initial diagnosis, n Stage III/Stage IV 76/21 43/7 33/14 0.036 58/16 18/5 0.471 Localized/generalized/MIP 20/71/6 10/34/6 10/37/0 16/52/6 4/19/0 Grade (A/ B/ C) 0/31/66 0/18/32 0/13/34 0.379 0/27/47 0/4/19 0.086 Duration, years 1.08 ± 0.66 0.97 ± 0.61 1.21 ± 0.69 0.070 0.10 ± 0.63 1.35 ± 0.67 0.023 SPC Adherence, n Regular SPC 55 30 25 0.499 40 15 0.345 Irregular SPC 42 20 22 34 8 Number 22.52 ± 9.16 21.80 ± 8.51 22.91 ± 9.96 0.554 20.95 ± 8.67 26.83 ± 9.65 0.007 Duration, years 10.17 ± 0.49 10.18 ± 0.50 10.17 ± 0.48 0.914 10.16 ± 0.49 10.20 ± 0.48 0.765 Mean BOP, % 16.81 ± 7.77 14.56 ± 5.28 18.16 ± 8.17 0.011 15.35 ± 6.20 19.37 ± 8.69 0.016 Mean GBI, % 5.97 ± 5.82 4.36 ± 3.30 7.13 ± 6.95 0.013 4.78 ± 4.09 8.67 ± 8.14 0.003 Mean PCR, % 31.10 ± 13.40 30.35 ± 14.45 32.23 ± 12.59 0.497 30.35 ± 14.33 34.17 ± 10.34 0.239 APT, active periodontal therapy; BOP, bleeding on probing; GBI, gingival bleeding index; MIP, molar-incisor pattern; n, number of patients; TL, overall tooth loss; PCR, plaque control record; SPC, supportive periodontal care; TLP, periodontal tooth loss. 3.2 Teeth

At T1, there were 2,323 teeth (1,074 [46%] anteriors, 679 [29%] premolars, and 571 [25%] molars). A total of 734 teeth (32%) were multi-rooted, of which 392 (53%) showed FI; 305 teeth (13%) showed tooth mobility, and 503 teeth (22%) were used as abutment teeth. Of the initial 2,360 teeth (T0), 2,224 teeth could be evaluated radiographically. BL at 136 teeth could not be assessed due to overlapping. A total of 2,118 teeth (90%) showed BL ≤60% and only 34 teeth (1%) >80%. Tooth-related data are shown in Table 2. Percentage as well as absolute frequencies of PD and CAL and mean GBI, plaque control record, and BOP are described in Table 3. Results of interindividual calibration have been reported before.7

TABLE 2. Tooth-specific characteristics Variables T0 T1 T2 P Teeth Number of teeth (per patient) 2,360 (24.33 ± 4.04) 2,323 (23.95 ± 4.20) 2,204 (22.72 ± 5.05) <0.0001 OTL during APT (per patient)

37

(0.38 ± 0.77)

OTL during SPC (per patient)

119

(1.23 ± 1.74)

TLP during SPC (per patient)

40

(0.41 ± 0.91)

Tooth type, n/% Anterior 1077/45.6 1074/46.2 1050/47.6 0.002 Premolar 686/29.1 679/29.2 643/29.2 <0.0001 Molar 597/25.3 570/24.6 511/23.2 <0.0001 Periodontal bone loss, n 0% to 20% 1180 n/a n/a 21% to 40% 662 n/a n/a 41% to 60% 276 n/a n/a 61% to 80% 72 n/a n/a >80% 34 n/a n/a Furcation involvement (FI) Single-rooted teeth, n 1,596 1,589 1,543 <0.0001 Multi-rooted teeth, n 764 734 661 <0.0001 Without FI, n/% 293/38.4 344/46.9 259/39.2 0.010 With FI, n/% 471/61.6 390/53.1 402/60.8 <0.0001 Degree I, n/% 256/54.3 254/64.8 283/70.4 0.620 Degree II, n/% 146/31.0 89/22.8 69/17.2 <0.0001 Degree III, n/% 69/14.7 47/12.4 50/12.4 0.004 Tooth mobility Without mobility, n 1777 2018 2128 <0.0001 With mobility, n 583 305 76 <0.0001 Degree I, n/% 417/71.5 251/82.3 63/82.9 <0.0001 Degree II, n/% 140/24.0 50/16.4 9/11.8 <0.0001 Degree III, n/% 26/4.5 4/1.3 4/5.3 <0.0001 Abutment status No abutment tooth, n 1856 1820 1679 <0.0001 Number of abutment teeth 506 503 525