A Comparison of Clinical and Radiological Parameters in the Evaluation of Molar Furcation Involvement in Periodontitis.

Objective: The aim of this study was to compare clinical, intra-surgical, 2D (panoramic) and 3D (CBCT)-based parameters in assessing molar furcation involvement (FI). Materials and
Methods: Six patients with generalized periodontitis Stage II to IV, Grade B and C who were scheduled for the periodontal flap surgical treatment were recruited in the study. In total, 38 molar teeth with 93 furcation sites were analysed. All subjects had comprehensive periodontal examination, which included an assessment of molar FI using Naber's probe according to modified Glickman's classification. Periodontal surgery was performed in patients with at least one maxillary molar with probing depth of ≥6 mm. This probing demonstrated lower grade of FI compared with intra-surgical findings.
Results: Periodontal probing, intra-surgical measurement and measurements based on CBCT significantly correlated with each other regarding the assessment of FI, with r values ranging between 0.81 to 1.00 (p<0.01). The correlation of panoramic radiograph with periodontal probing is 0.49, with CBCT 0.39 and with intra-surgical measurements 0.36. The results showed an excellent agreement and higher accuracy between intra-surgical measurements and CBCT (0.96), in contrast to clinical examination and panoramic radiography- 0.87 and 0.63 respectively. Different clinical and radiological modalities showed a correlation among each other. They are accurate and have their own benefits, which makes (renders) them useful in establishing periodontal diagnosis and treatment planning.
Conclusion: However, CBCT offers significant advantages including excellent agreement and higher accuracy and can be used as justified as excellent diagnostic tool in detecting and locating FI to provide a more reliable diagnosis and basis for treatment decisions.

Introduction

According to American Academy of Periodontology, a furcation involvement is a condition in which periodontitis is affecting root trunk causing resorption of bone into the bi- or trifurcation area of a multi-rooted tooth (). The extent of the defect and the position of marginal tissue in its relationship to the FI provide relevant clinical information for more accurate and reliable diagnosis and predictable prognosis, by helping to select an adequate treatment option. Furcation involvement is among other factors (clinical attachment loss, probing depth, tooth mobility) an important guide in clinical prognosis and therapeutic decisions (). Therapy of FI is many times based on clinically identified extent of the involvement (). The grade of FI contributes significantly to the prognosis of the tooth at following periodontal treatment and during the maintenance therapy (). The treatment of the cases with FI presents a great challenge since the furcation is an area of specific anatomic morphology that may be challenging to debride with standard periodontal instrumentation (-) and are difficult to reach by standard oral hygiene measures. 81% of furcations have diameter entrances <1,00mm and 58%<0,75mm (). Narrow entrances, presence of ridges, convexities and concavities often do not allow for adequate instrumentation or complete plaque removal by patient (). The scaling and root planing in the furcation area proved to be more effective when the treatment is performed surgically, emphasizing the importance of adequate clinical diagnosis (, ). However, physical access, morphological variations and measurement errors can limit the correct assessment of FI (-), thus leading to alteration of treatment plan and adding unanticipated treatment costs (, ).

Traditionally, the FI is assessed by utilizing both clinical and radiographic examinations. Clinical examination is performed with curved scaled Naber’s probe and FI is categorized according to the one of the proposed classification systems. To describe the extent and features of the furcation defect, 1953. Glickman developed one of the first classification system, which is still today one of the most widely used classification systems for assessing FI ():

Grade I is the incipient or early involvement. There is a supra-alveolar pocket with minimal bone loss in the furcation area. A radiographic change is not detectable, while the probe penetration is clinically absent.

Grade II includes cases with bone destruction on one or more aspects of the furcation. However, some portions of the soft and hard tissues remain intact, which does not allow complete penetration of the probe into the furcation area. Radiographic changes may or may not be present.

Grade III means that the inter-radicular bone is not present, but the orifices of furcations are covered with soft tissue only. The furcation entrance is not clinically visible, although there is a through- and- through- lesion present. Usually, this destruction will appear as a radiolucency between the roots, especially in the mandible.

Grade IV is the most severe involvement with a complete destruction of the inter-radicular bone in the furcation area. The furcation opening is clinically visible because the soft tissues also receded more apically. The radiographic image basically corresponds to the grade III findings. The imaging of periodontal structures complements the clinical examination and can be achieved through extraoral panoramic radiographs, intraoral periapical and bite-wing radiographs (). However, 2D radiographs are limited due to projection geometry, since they allow only two-dimensional presentation of three-dimensional periodontal structures.

The standardized clinical protocol used in diagnostics of periodontitis was described for the first time more than 50 years ago and has not changed much since then (), The advancements in dento-alveolar imaging could overcome the limitations of conventional intraoral radiographs providing the true three-dimensional imaging by using the cone beam computed tomography (CBCT). CBCT generates 3D volumetric images and it has been commonly used in dentistry. All CBCT units provide multi-planar axial, coronal and sagittal reconstructed images without magnification (). In contrast to 2D intraoral radiographs and periodontal probing, 3D CBCT imaging was found to be more effective in evaluating periodontal structures. CBCT images have shown better potential for detecting periodontal bone defects in all directions compared with periapical radiographs (). CBCT imaging is as accurate as clinical measurements with a periodontal probe and as reliable as intraoral radiographs for interproximal areas. Considering all numerous advantages which CBCT offers to provide accurate diagnosis, it is currently being considered as a superior diagnostic tool for various applications in periodontology (). Due to high accuracy and various advantages (low radiation exposure, rapidity of scan time, reduced equipment costs), CBCT provides benefits in periodontal diagnostics, especially for advanced and complex periodontal disease including detailed information regarding the amount of bone loss, involvement of furcation, type of defects and their dimensions. Additionally, it is helpful in determining more accurate prognosis for each tooth by allowing 3D analysis of the surrounding bone, which influences the decision making process in periodontology, especially for periodontal regeneration procedures (, ). According to current available evidence, American Academy of Periodontology has stated that a 2D full-mouth radiograph combined with clinical periodontal probing remain the gold standard for a comprehensive evaluation of periodontal structures (). However, experts on the best evidence consensus panel identified several scenarios where addition of CBCT imaging would be useful, including the scenario in which an advanced FI has been diagnosed (). Although there are several in vivo and in vitro studies in assessing the accuracy of CBCT in the measurement of periodontal bone defects, there are just two studies that have investigated the efficacy of CBCT in diagnosing FI by comparing its result to intra-surgical assessment (, ). In both studies CBCT images demonstrated a high accuracy in evaluating the loss of periodontal tissue and classifying the degree of FI. The aim of this study was to evaluate (compare and correlate) clinical, intra-surgical, 2D (panoramic) und 3D (CBCT)-based parameters in assessing molar FI.

Subjects and Methods

Six patients with generalized periodontitis Stage II to IV, Grade B and C who were scheduled for the periodontal flap surgical treatment were recruited in the study. The study was approved by the Ethics Committee of University of Zagreb. Inclusive criteria were: completion of initial periodontal non-surgical therapy, the presence of at least 15 teeth, at least 2 interproximal areas with a loss of attachment of ≥4 mm or at least 2 interproximal areas with pockets depth ≥5 mm, but not on the same tooth, criteria for periodontitis in at least 30% of the teeth present (generalized periodontitis), optimal oral hygiene, FMPS and FMBS <20%. The initial therapy, including oral hygiene instruction and motivation, scaling and root planing and occlusal adjustment were performed and re-evaluations were scheduled following 3 and 6 months. After completion of initial non-surgical periodontal therapy, a written consent was taken from the patients prior to participation in the study. Periodontal surgery was performed only in patients with at least one maxillary molar with probing pocket depth of ≥6 mm.

All subjects underwent a comprehensive periodontal evaluation, which included an assessment of molar FI using Naber’s probe according to modified Glickman’s classification (): Class I, horizontal bone loss < 2 mm into the furcation; Class II, horizontal bone loss deeper than 2 mm but less than 6 mm into the furcation; Class III, extensive horizontal bone loss with a through-and-through lesion.

The panoramic dental radiographs were taken using Sirona Orthophos XG X-ray unit (Sirona Dental Systems GmbH, Fabrikstrasse 5, 64625 Bensheim, Germany) set at 64 kV and 8 mA with exposition time 14,1 s. The presence of triangular radiolucency at the furcation area was radiographic sign for FI, which was recorded as present or absent.

CBCT scans were obtained with Planmeca ProMax 3D CBCT (Planmeca Oy, Asentajankatu 6, 00880 Helsinki, Finland) with 90kV, 10mA, FOV 1001x1001x999mm, 360º rotation, exposition time 18,071s, voxel size 200 µm. Third quartile of dose area product (DAP) was 1555.9 mGy x cm2. CBCT images were obtained using Romexis Viewer Planmeca 3.8.3.R. (Planmeca Oy, Asentajankatu 6, 00880 Helsinki, Finland), generated in the digital imaging and communications in medicine (DICOM) format and analyzed by axial and sagittal reconstructions with cutting interval of 1mm. The FI was presented as a trabecular bone resorption at the furcation area on both sagittal and axial view. The depth was determined on the axial slice as a distance from a line that was drawn tangentially to the neighboring root surfaces to the deepest point of bone loss.

After the CBCT scan, periodontal flap surgery was performed. Following administration of local anesthesia, a full thickness mucoperiostal flap was raised. Direct clinical intra-surgical measurements were made prior to complete scaling and root planing. Since the measurements in Glickman’s classification are made in presence of soft tissues, the intra-surgical FI assessments were performed according to modified Glickman’s classification (). FI intra-surgically and on CBCT were assessed at three sites (buccal, mesio-palatal and disto-palatal) of maxillary molars and two sites (buccal and lingual) of mandibular molars using Naber’s probe. In order to eliminate inter-examiner discrepancies the same investigator performed all clinical and radiological measurements in all patients.

True-positives (TP - the number of cases correctly identified as FI), false-positives (FP - the number of cases incorrectly identified as FI), true-negatives (TN- the number of cases correctly identified as absence of FI) and false-negatives (FN - the number of cases incorrectly identified as absence of FI) were determined using intra-surgical findings as the gold standard. This evaluation of FI was assessed using dichotomous scale (present/absent) and according to modified Glickman's scale. Test characteristics were calculated (sensitivity, specificity, positive predictive value -PPV, negative predictive value-NPV, diagnostic accuracy -ACC, false discovery rate -FDV, diagnostic odds ratio –DOR) using following formulas:

The sample size was calculated using power analysis before the initiation of the study, assuming a difference of <5% between radiographic and surgical measurements and using the formula (zα)2 × (s)2/(d)2. The collected data were subjected to statistical analysis using SPSS software for Windows (SPSS Inc, Chicago, IL). The Kappa statistics was used to determine intra-observer agreement. The kappa values were interpreted as recommended by Landis and Koch and adapted by Altman: k≤0.20 poor, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 good, 0.81-1.00 very good (). The Pearson’s correlation test was used to correlate the measurements performed clinically and radiologically. The statistical significance was set at p<0.05. Accuracy, sensitivity, specificity, positive and negative predictive values and accuracy were calculated with the McNemar χ2 test.

Results

The study was conducted on 6 patients: 2 were females, 4 males and an age range of 35 -77 with a mean age 53.50±14.80. In total, 38 molar teeth with 93 furcation sites were analysed (9 maxillary first molars, 8 maxillary second molars, 10 mandibular first molars, 11 mandibular second molars). The kappa values for intra-observer agreement ranged between good and very good, as follows: 0.72 for periodontal probing, 0.81 for intra-surgical measurements and 0.85 for CBCT measurements.

The comparison of maxillary molar assessment showed that probing generally demonstrated lower grade of FI compared with intra-surgical findings. For 5,88% of cases periodontal probing could not detect bone loss in class III furcations and also failed to detect any bone loss in class I furcations. On the contrary, CBCT showed high agreement with intra-surgical examination. The maxillary molar FI assessed by periodontal probing, panoramic radiograph and CBCT is illustrated in Table 1. A comparison of mandibular furcations revealed scenarios in which no FI was detected clinically, however 19,04% of cases were demonstrated to have some class of furcation involvement by intra-surgical findings. Similarly to maxillary molars, CBCT demonstrated high correlation with intra-surgical examination. The mandibular molar FI evaluated by clinical examination, intra-surgical and CBCT interpretation is shown in Table 2.

Table 1

Maxillary molar FI of evaluated sites assessed by periodontal probing, intrasurgical measurement and measured by CBCT.

	Buccal			Mesial palatal			Distal palatal
Modified Glickman'sClassification	Periodontal probing	Intrasurgical measurement	CBCT	Periodontal probing	Intra-surgical measurement	CBCT	Periodontal probing	Intra-surgical measurement	CBCT
Not present	58.82%	52.94%	52.94%	52.38%	58.82%	58.82%	64.71%	52.94%	58.82%
Class I	35.29%	29.41%	29.41%	23.81%	29.41%	29.41%	29.41%	35.29%	29.41%
Class II	5.88%	11.76%	11.76%	4.76%	5.88%	5.88%	5.88%	5.88%	5.88%
Class III	0	5.88%	5.88%	0	5.88%	5.88%	0	5.88%	5.88%

Table 2

Mandibular molar FI of evaluated sites assessed by periodontal probing, intrasurgical measurement and measured by CBCT.

	Buccal			Lingual
Modified Glickman'sClassification	Periodontal probing	Intra-surgical measurement	CBCT	Periodontal probing	Intra-surgical measurement	CBCT
Not present	61.90%	42.86%	42.86%	61.90%	42.86%	47.62%
Class I	23.81%	38.10%	38.10%	28.57%	42.86%	38.10%
Class II	14.29%	9.52%	14.29%	9.52%	9.52%	4.76%
Class III	0	9.52%	4.76%	0	4.76%	9.52%

The Spearman’s correlation demonstrated that periodontal probing, intra-surgical measurement and measurements based on CBCT significantly correlated with each other in the assessment of FI, with r values ranged between 0.81 to 1.00 (p<0.01; Table 3). The largest agreement (100%) was found in buccal maxillary sites between CBCT and intra-surgical measurement. The smallest agreement (81%) was found in lingual mandibular molars, in which 19% of FI was detected using CBCT, although not clinically. The correlation of panoramic radiograph with periodontal probing was 0.49, with CBCT 0.39, and with intra-surgical measurements it was 0.36.

Table 3

Correlation coefficients of periodontal probing and CBCT-measurements with intrasurgical measurements in assessment of FI (correlation is significant at p<0.01).

	Periodontal probing/intra-surgical	CBCT/intra-surgical	Periodontal probing/CBCT
Maxillary buccal	0.91	1.00	0.91
Maxillary mesial palatal	0.92	0.92	0.92
Maxillary distal palatal	0.88	0.96	0.92
Mandibular buccal	0.87	0.98	0.85
Mandibular lingual	0.83	0.94	0.81

The results showed an excellent agreement and higher accuracy between intra-surgical measurements and CBCT (0.96), in contrast to clinical examination and panoramic radiography- 0.87 and 0.63 respectively (Table 4). CBCT can be used as highly sensitive (0.93) and specific test (1.00) to accurately identify the FI when it is really present and to rule out the FI if it is not radiologically detectable. On the other hand, panoramic radiographs showed low sensitivity and high specificity, which means that these tests give a few false positive results but they are unable to identify the majority of the positive cases of FI (69%). Periodontal probing showed generally satisfying sensitivity (0.74) and high specificity (1.00). Precision of periodontal probing and CBCT was higher (PPV=1.00) compared to panoramic radiographs (PPV=0.88). The false discovery rate and diagnostic odds ratio as useful indicators of test performance were not comparable, since the values by periodontal probing and CBCT were 0, which can be attributed to a small sample size.

Table 4

Sensitivity, specificity, positive predictive value, negative predictive value, accuracy, false discovery rate and diagnostic odds ratio in detection of FI by periodontal probing, panoramic radiograph and CBCT using intra-surgical findings as the gold standard.

	Sensitivity	Specificity	Positive predictive value(PPV)	Negative predictive value(NPV)	Accuracy(ACC)	False discovery rate (FDR)	DOR (diagnostic odds ratio)
Periodontal probing	0.74	1.00	1.00	0.79	0.87	0	0
Panoramic radiograph (OPG)	0.31	0.95	0.88	0.57	0.63	0.13	8.53
CBCT	0.93	1.00	1.00	0.93	0.96	0	0

Discussion

Different diagnostic methods showed significant correlation among each other and the results confirmed the clinical relevance of CBCT in the FI assessment, since CBCT showed a strong agreement with the direct intra-surgical findings (the current “gold standard”) in the detection of FI. Although all included patients were diagnosed with generalized periodontitis, more than half of them showed no FI based on the four evaluation methods (52.94-64.71%).

Accuracy of clinical detection of FI is unpredictable, since it depends on many factors, such as: operator technique and experience (probe angulation, amount of force exerted, access), tooth position, inclination, presence of adjacent teeth, length of root trunk, root morphology, roots divergence and configuration of residual inter-radicular bone (, ). The results of our study showed that periodontal probing of maxillary and mandibular molars generally underestimated the extent of FI, suggesting that clinical detection is unreliable and should be supplemented with radiographs (, ). In many cases, the clinical measurement reflects the probing depth into the inflamed connective tissue, instead of the actual depth of the inter-radicular bony defect (). The study of Graetz et al. demonstrated that determining the degree of FI by clinical probing was accurate in only 56% of assessed cases (). In our study, the clinical detection identified the absence of FI, although the 5.88-11.77% of cases demonstrated bone loss intra-surgically and on CBCT images, thus indicating under-detection. Periodontal probing also failed in detecting class III furcations. On the contrary, clinical examination showed over-detection of class I furcations. The results are in accordance with findings from other studies that showed that clinical detection can easily lead to under- or over-estimation of FI (, ). The findings confirmed the necessity of supplementing clinical detection with radiographic examination, which is in accordance with the consensus in the literature (, ).

The results of our study showed that the correlation of panoramic radiograph (orthopantomogram) with other detection methods was low (0.36-0.49). Precision of panoramic radiograph was satisfactory (PPV=0.88). However, panoramic radiograph had low sensitivity (0.31) and high specificity (0.95) for FI detection, mainly due to intrinsic limitations of 2D imaging such as anatomic complexity (superimposition of palatal root at the furcation region, sinus tract extending into furcation) and angulation problems (, -). The detectability of early stages of FI on 2D imaging is especially inconsistent and limited ().

The correlation of CBCT with intra-surgical findings was very high: 0.92-1.00 for maxillary molars and 0.94-0.98 for mandibular molars. CBCT was able to detect significant bone loss in class III furcations (more than 6mm), when other methods failed. Sensitivity and negative predictive value of CBCT in FI detection was 0.93 and specificity and positive predictive value 1.00, which suggests that CBST is a valuable tool in detection of FI, thus offering significant advantage over conventional clinical and radiographic assessments. Although the direct intra-surgical exploration is the most accurate way of assessing the degree of FI (), the invasiveness and difficulty of performing it, make this method often inapplicable. CBCT reveals precisely and accurately the alveolar bone resorption, infra-bony pockets and furcation defects () and since the accuracy in our study was 0.96, it can be concluded that CBCT is capable to generate reliable and precise radiographs in patients with generalized periodontitis making it an excellent adjunctive diagnostic tool in periodontal treatment planning, which is in accordance with findings from other authors (, , , -). The studies of Walter and Qiao showed that the 84% and 82.4% of CBCT data, respectively, were confirmed by intra-surgical findings and assessment of maxillary molar FI (38%). The differences may be explained by the surgical protocol probably leading to a minor loss of periodontal tissues during instrumentation, and that is the reason why we performed the measurements prior to scaling and root planing with debridement. A CBCT analysis was performed on hard tissue defects only, while intra-surgical measurements did not include flap thickness in the furcation area. On the other hand, periodontal probing of horizontal FI was measuring the supracrestal attached tissue. Therefore, different measuring methods might explain a lower correlation of clinical and intra-surgical or CBCT measurements. Another explanation for errors that exist between CBCT and direct surgical measurements can lie in the fact that there were different accuracies of those measurements. Clinical measurements were only able to be performed to the nearest 0.5mm, whereas CBCT measurements can be made to the nearest 0.1mm. Walter suggested that the additional CBCT provided not only detailed information of FI but also facilitated a clear decision for additional periodontal treatment, when compared to treatment recommendations from clinical findings and 2D (periapical) imaging (). Furthermore, CBCT allows a reduction in treatment costs and time for periodontally compromised maxillary molars (). Although the full-mouth radiographic series are considered to be current standard in a periodontal diagnostics, especially due to their orthoradial projection, in our study we compared the orthopantomograms since they were already present as part of first examination of the patients. Additional patient irradiation with both CBCT and full-mouth radiograph was not performed due to ethical issues. Orthopantomogram is often already available since it is performed as part of general dental screening of patients. Furthermore, there is the lack of studies that examined orthopantomograms in a comprehensive periodontal diagnostics. The application of CBCT as a promising tool with superior image quality is growing rapidly in dentistry including periodontology. Although it is considered a valuable addition to periodontal clinical assessment, CBCT is not without shortcomings and limitations. Noise, scatter, patients related artefacts, partial volume averaging and beam hardening artefacts (cupping and streak artefacts) could compromise its diagnostic quality, especially for patients with heavy metallic restorations, orthodontic appliances, multiple endodontic treatment or implants (, ). Its effective radiation dose is still 1,8 times higher than in conventional panoramic radiography and it exposes the sensitive tissues in head and neck region to radiation (, ). Since the dose varies depending on the device, the field of view and factors of the selected technique, a dose should be reduced by using smaller volume in the region of interest consistent with clinical indications (). The clinical examination and conventional 2D imaging should continue to be used as routine examination in periodontal assessment. At this time, a routine use of CBCT for the diagnosis and treatment of moderate-to-severe periodontitis does not appear to be warranted from radiation exposure and cost perspective (). However, in selective cases a limited view of CBCT may be useful and should be considered as an adjunctive diagnostic method after a comprehensive periodontal examination as a useful and widely available tool that has the potential to improve today’s standard of care by providing some significant changes in the course of treatment (, ). For complicated cases when standard examination fails to provide sufficient information for diagnosis and/or treatment planning, CBCT may be added with the smallest available field of view and optimally selected exposure settings ().

Conclusion

Our study suggests that different clinical and radiological modalities show a correlation among each other. They are satisfyingly accurate and have benefits, which makes them useful in establishing periodontal diagnosis and aid in treatment planning. However, CBCT offers significant advantages including excellent agreement and higher accuracy; therefore, it can be used as justified and excellent diagnostic tool in detecting and locating FI and a reliable basis for treatment decisions. Its application should be considered carefully through precise indication in relation to its limitations and risks.