Coronal Restoration as a Predictor of Periapical Disease in Non-Endodontically Treated Teeth.

OBJECTIVES: To compare the frequency of apical periodontitis (AP) in non-endodontically treated teeth restored with full coverage crowns, amalgam and resin composite fillings, and to disclose the association of type and quality of coronal restoration with periapical disease.
MATERIAL AND METHODS: The cross-sectional study involved 597 subjects who attended the Dental Clinic of the Clinical Hospital Centre, Rijeka, Croatia for the first time. Data were collected by means of a clinical and radiological survey. The following data were recorded for all permanently restored non-endodontically treated teeth: the type, the quality of coronal restoration according to clinical and radiographic criteria for marginal integrity as well as the recurrent caries presence, periapical status and marginal bone loss. Chi-square tests were used to analyze the difference in periapical status regarding the type and the quality of restoration. The multivariate logistic regression analysis was used to assess the effect of explanatory tooth-specific variables on periapical status.
RESULTS: There was no significant difference in the proportion of AP with regards to the type and material of adequate quality coronal restorations (χ2=5.000, P=0.082). The regression analysis revealed significant positive associations with outcome variable (AP) for crowns (OR=3.39; P<0.001), recurrent caries (OR=4.67; P<0.001) and premolar teeth (OR=1.73; P=0.002).
CONCLUSION: If the quality of restorations is adequate, there is no difference in the periapical status of non-endodontically treated teeth regarding the type and material of coronal restoration.

Introduction

Apical periodontitis (AP) is a common inflammatory disease affecting periapical tissues due to microbial infection of root canals. The incidence of AP varies greatly among countries considering the differences in sampling methods and diagnostic criteria (). A study on the prevalence of AP in Croatian adults reported the presence of periapical disease in 8.5% of the examined teeth (). Several risk indicators specific for the occurrence of AP have been identified. A significantly increased risk for the presence of AP was reported for teeth restored with coronal fillings or crowns, teeth with primary caries, endodontically treated teeth, and molars (). A longitudinal epidemiologic study reported an association of periapically diseased non-endodontically treated teeth with the presence of coronal restoration of poor quality or carious lesion (). However, limited data on the association of AP and type of coronal restoration in non-endodontically treated teeth are available.

Despite recommendations for a global phase-down, amalgam is still widely applied material for restoration of posterior teeth in certain countries due to its low cost and effectiveness (Minamata Convention on Mercury). In Croatia, amalgam is mostly utilized in dental settings because it provides dental care through the health insurance system. Alternatively, resin composite restorations require highly demanding adhesive procedure that can be challenging in oral environment. Furthermore, polymerization shrinkage, stress, debonding and marginal leakage as the main causes of adhesive restorations failure can lead to pulpal breakdown, and subsequently to AP (). Low-quality evidence suggested a higher failure rate in resin composite restorations and an increased risk of recurrent caries in contrast to amalgam restorations (). An in vitro study found that microleakage was significantly higher in composite than amalgam restorations (). A cross-sectional study conducted in Sweden hypothesized that teeth restored with composite resin restorations exhibit AP more frequently than teeth restored with amalgam restorations, yet no significant differences in the frequency of periapical pathology regarding these materials were found ().

Several studies reported higher prevalence rates of AP in teeth restored with crowns (, ). However, long-term follow-up studies reported low incidence of radiographically detectable periapical lesions in crowned teeth (, ). A study investigating factors related to pulpal breakdown in vital teeth restored with metal-ceramic crowns suggested a significantly higher survival of the pulp vitality in single crowned teeth than in teeth that served as abutments of a fixed bridge ().

The hypothesis for the present study was that the periapical status of non-endodontically treated teeth does not vary with regards to the type of permanent coronal restorations. The purpose was to compare the frequency of AP in non-endodontically treated teeth restored with full coverage crowns, amalgam and resin composite fillings, as well as to disclose the association of type and quality of coronal restoration with the presence of AP.

Material and methods

This cross-sectional study received ethical approval from the Institutional Ethical Committee (003-05/13-01/03). The sample that formed the basis for the study consisted of 1072 patients older than 18 years, who attended Dental Clinic of the Clinical Hospital Centre, Rijeka, Croatia for the first time and who presented consecutively within two years. The patients were not included if they refused to participate, had seven or less remaining teeth, received endodontic therapy within two years, and were unwilling or unable to attend the radiographic diagnostics. The implementation of these criteria provided the sample of 597 subjects (Figure 1). All participants accepted to take part by signing an informed consent. The study was conducted in accordance with the World Medical Association Declaration of Helsinki principles.

Figure 1

Flowchart demonstrating patient enrolment in the study

Data were collected by means of a clinical and radiographic survey. The radiographic survey included digital panoramic radiographs and 4 bitewings (revealing the distal surface of canine to mesial surface of the last molar). Panoramic images were obtained by utilizing a panoramic digital radiography device (J. Morita Corporation, Veraviewepocs 6716, Kyoto, Japan) with exposure parameters following the manufacturer's recommendation depending on the patient's size and weight (70 - 90 kV, 10-15 mA, and 14 s of exposure). Images were processed using software (Mediadent V4, Image Level, Nieuwkerkenwaas, Belgium). Bitewings were taken using an X‐ray unit (Trophy Elitys, Trophy Radiologie, Marne‐la‐Vallee, France) and an intraoral sensor (One, Owandy Radiology, Roslyn, NY, USA). The exposure parameters used were 60 kV, 7 mA and 0.32 s. Panoramic and bitewing images were analyzed on a 19-inch liquid crystal monitor (P1914S; Dell, Austin, TX, USA; resolution:1,280 x 1,024 32-bit color; graphic card: HD Graphic; Intel, Santa Clara, CA, USA).

The following data were recorded for all present teeth, apart from third molars and impacted teeth: coronal status (intact tooth, primary caries, temporary restoration, permanent restoration), the presence of endodontic treatment and periapical status. For diagnosis of dental caries, the WHO diagnostic thresholds were used (World Health Organization 1997; 13). The total number of teeth present in the sample for this survey was 14852. The analyses of AP presence regarding the quality and type of coronal restoration were performed on 5298 permanently restored non-endodontically treated teeth. All endodontically treated teeth and 8063 non-endodontically treated teeth (intact teeth, teeth with primary caries or temporary restoration) were omitted from further analysis (Figure 2).

Figure 2

Flowchart demonstrating teeth enrolment in the analysis of apical periodontitis presence regarding the type of restoration

Assessment of restorations

Using the recordings from clinical examination and interpretation of radiographic images (panoramic radiographs and bitewings), following data for all non-endodontically treated teeth with permanent restorations were noted: type (filling or crown), material (amalgam or resin composite) and quality of restoration according to clinical and radiological criteria for marginal integrity and recurrent caries presence. The clinical quality of coronal restorations was evaluated in accordance with modified United States Public Health Service/Ryge (USPHS/Ryge) criteria for marginal integrity and recurrent caries as reported by Merdad et al. (). Radiographically, marginal integrity of the restoration was defined as adequate (radiographically sealed) or inadequate (signs of open margins or overhangs) according to previously described criteria (). Recurrent caries was noted as absent or present (clearly visible reduction in mineral content of a proximal tooth surface contiguous with a restoration; ().

Assessment of periapical status and marginal bone loss

Analysis of the marginal bone loss and periapical status was accomplished using digital panoramic radiographs. The loss of the marginal bone was designated as less or more than one third of the root length (). The periapical status was analyzed using the periapical index system (PAI; 17). Each tooth was scored using visual references for the full-scale PAI. To define the periapical status in multirooted teeth, the highest PAI value of all roots was used. The full-scale scores were dichotomized, and periapical status was registered as healthy (PAI scores 1 and 2) or AP (PAI scores 3, 4 and 5).

Calibration procedures

The PAI calibration was accomplished by evaluation of periapical status in 100 reference teeth (). After each tooth was given one of the five PAI scores, a comparison to gold standard scores was made, and a Cohen´s Kappa was calculated (0.70). Calibrations for diagnosis of dental caries, coronal restoration quality and marginal bone loss were performed according to WHO recommendations (). Diagnostic intra-examiner agreement was analyzed by double scoring of the 30 randomly selected individuals´ clinical and radiographic surveys. Intra-examiner agreement was determined by calculating Cohen's Kappa for clinical and radiographic diagnosis of recurrent caries (0.85 and 0.92, respectively), clinical and radiographic assessment of marginal integrity (0.81 and 0.85, respectively), PAI (0.75) and marginal bone loss (0.89).

Statistical analysis

For statistical analysis, a software (Statistica 13.0, Statsoft, Tulsa, OK, USA) was used, at level of statistical significance P<0.05. To test data for distribution normality Lilliefors test was used. As the distribution of data was not normal, median and interquartile range were used to report central tendency and dispersion. Mann-Whitney U test was used to test the differences between the groups regarding the continuous variables. Chi-square tests were used to analyze the variations in periapical status regarding the type and quality of restoration as well as tooth type. Multivariate logistic regression analysis was used to assess the effect of explanatory tooth-specific variables (type of restoration, marginal integrity, recurrent caries, tooth type and marginal bone loss) on periapical status. All explanatory variables are provided in table 1. The outcome variable was defined as the presence versus absence of AP in the tooth.

Table 1

Independent variables used in multivariate logistic regression analysis

Type of restoration
Amalgam filling
Composite resin filling
Crown
Marginal integrity of restoration (clinically and radiographically)
Adequate	No defective restoration margin
Inadequate	Defective restoration margin
Recurrent caries (clinically and radiographically)
No	No detectable caries contiguous with a restoration
Yes	Detectable caries contiguous with restoration
Tooth typeIncisors and caninesPremolarsMolars
Marginal bone loss (radiographically)
No marginal bone loss	<1/3 root length
Marginal bone loss	>1/3 root length

*First category was used as reference category

Results

The final sample involved 190 male (31.8%) and 407 female participants (68.2%). The median age of participant was 34 years (interquartile range 24 - 46). No difference in age was detected regarding participants' sex (Mann-Whiney U test; P=0.534). The median number of remaining teeth ranged from 27 in the youngest age group to 22 in the oldest age group. On the average, an examined individual had 1 non-endodontically treated tooth with AP (interquartile range 0-1). The presence of AP in 1or more non-endodontically treated teeth was recorded in 54.3% of the examined participants (Table 2).

Table 2

Distribution of all examined individuals regarding age, dental and periapical status

Age	Examined individuals, n	Remaining teeth (median, interquartile range)	Intact teeth(median, interquartile range)	Non-endodontically treated teeth with AP(median, interquartile range)	Individuals with 1 or more non-endodontically treated teeth with AP, n (%)
<29	244	27 (26-28)	16 (12-18)	1 (0-1)	110 (45.1)
30-39	129	27 (25-28)	14 (10-16)	0 (0-1)	70 (54.3)
40-49	105	25 (24-26)	10 (9-14)	0 (0-1)	70 (66.7)
50-59	69	24 (21-24)	8 (4-10)	1 (0-2)	39 (56.5)
>60	50	22 (13-23)	7 (2-10)	1 (0-2)	35 (70.0)
Total	597	26 (24-28)	12 (9-16)	1 (0-1)	324 (54.3)

AP-apical periodontitis

The percentages are calculated out of examined patients in each age group.

AP was diagnosed in 286 (5.4%) permanently restored non-endodontically treated teeth. The frequency of AP in relation to the type and quality of coronal restorations is demonstrated in Table 3. A significant difference in the prevalence of teeth with AP according to the type of coronal restoration was present (χ2=35.344, P<0.001). The frequency of AP in teeth restored with crowns was significantly higher than in the teeth restored with amalgam (12.0% vs 5.1%; P<0.001) or those restored with composite fillings (12.0% vs 4.8%; P<0.001). If the overall quality of restoration was adequate, there were no significant differences in the proportion of AP with regards to type of coronal restoration (χ2=5.000, P=0.082; Table 3).

Table 3

Type and quality of coronal restoration in relation to AP

	AP in relation to restoration			AP in relation to restoration quality
				Adequate quality			Inadequate quality
Type/material of restoration	Total	AP, n (%)	Statistics	Total	AP, n (%)	Statistics	Total	AP, n (%)	Statistics
Amalgam	1363	70 (5.1)	χ2=35.344P<0.001*	768	14 (1.8)	χ2=5.000P=0.082†	595	56 (9.4)	χ2=44.953P<0.001*
Composite	3551	170 (4.8)		1912	40 (2.1)		1639	130 (7.9)
Crown	384	46 (12.0) ‡		239	10 (4.2)		145	36 (24.8) ‡
Total	5298	286 (5.4)		2919	64 (2.2)		2379	222 (9.3)

*Significant differences in frequency of teeth with AP according to type of coronal restoration.

†No significant difference in frequency of teeth with AP according to type of coronal restoration.

‡Significant differences vs both other types of coronal restoration

Table 4 demonstrates the distribution of teeth with AP in relation to the tooth group, quality and type of coronal restoration. Significant differences in the proportion of teeth diagnosed with AP regardless of the quality of coronal restoration were detected in the group of premolars (χ2=12.872; P=0.002) and molars (χ2=34.468; P<0.001). In the premolar group, AP was more commonly detected in crowned teeth than in teeth restored with amalgam (13.4 vs 4.3; P=0.002) or composite fillings (13.4 vs 5.9; P=0.003). Molars restored with crowns were also more often diagnosed with AP than molars restored with amalgam (19.0 vs 5.4; P<0.001) or composite fillings (19.0 vs 4.1; P<0.001). However, only the group of premolars demonstrated significant differences regarding the type of coronal restoration (χ2=19.786; P<0.001) in teeth with adequate quality of coronal restoration. AP was more common in premolars restored with crowns, than those restored with amalgam or composite fillings (11.9% vs 1.9% and 3%, respectively; both P<0.001; Table 4).

Table 4

Type and quality of coronal restoration in relation to AP and tooth type

	AP in relation to type/material of restoration			AP in relation to restoration quality
				Adequate quality			Inadequate quality
	Total	AP, n (%)	Statistics	Total	AP, n (%)	Statistics	Total	AP, n (%)	Statistics
Incisors and canines
Amalgam	5	0 (0)	χ2=4.777P=0.092†	5	0 (0)	χ2=4.003P=0.135†	0	0 (0)	χ2=23.388P<0.001*
Composite	1024	45 (4.4)		484	15 (3.1)		540	30 (5.6)
Crown	186	15 (8.1)		121	0 (0)		65	15 (23.1)
Total	1215	60 (4.9)		610	15 (2.5)		605	45 (7.4)
Premolars
Amalgam	323	14 (4.3)	χ2=12.872P=0.002*	213	4 (1.9)	χ2=19.786P<0.001*	110	10 (9.1)	χ2=2.046P=0.360†
Composite	1179	70 (5.9)		674	20 (3.0)		505	50 (9.9)
Crown	119	16 (13.4) ‡		84	10 (11.9) ‡		35	6 (17.1)
Total	1621	100 (6.2)		971	34 (3.5)		650	66 (10.2)
Molars
Amalgam	1035	56 (5.4)	χ2=34.468P<0.001*	550	10 (1.8)	χ2=4.223P=0.121†	485	46 (9.5)	χ2=29.324P<0.001*
Composite	1348	55 (4.1)		745	5 (0.7)		594	50 (8.4)
Crown	79	15 (19.0) ‡		34	0 (0)		45	15 (33.3) ‡
Total	2462	126 (5.1)		1338	15 (1.1)		1124	111 (9.9)

*Significant differences in frequency of teeth with AP according to type of coronal restoration.

†No significant difference in frequency of teeth with AP according to type of coronal restoration.

‡Significant differences vs both other types of coronal restoration

Table 5 demonstrates the results of a multivariate logistic regression analysis with all considered explanatory variables included in the model. A significant positive association was found between the outcome variable, AP and crowns (OR=3.39; 95% CI 2.2-5.3; P<0.001), recurrent caries (OR=4.67; 95% CI 2.8 to 7.8; P<0.001) and premolars (OR=1.73; 95%CI 1.2-2.4; P=0.002; Table 4). The highest association was determined for the presence of recurrent caries. Conversely, it was found that teeth with inadequate marginal integrity of restorations have a slightly increased, but not statistically significant risk for AP presence than teeth with an adequate marginal integrity. Marginal bone loss >1/3 root length was not statistically associated with an increased risk of AP presence (Table 5).

Table 5

Multivariate logistic regression analysis of tooth-specific variables in relation to AP in restored teeth

Variables	Odds ratio	95% CI	P value
Type of restoration
Amalgam	1	Reference
Composite	0.918	0.672 to 1.255	0.593
Crown	3.390	2.181 to 5.269	<0.001
Marginal integrity of tooth restoration
Adequate	1	Reference
Inadequate	1.359	0.780 to 2.368	0.279
Recurrent caries
No	1	Reference
Yes	4.666	2.769 to 7.862	<0.001
Tooth type
Incisors and canines	1	Reference
Premolars	1.725	1.217 to 2.446	0.002
Molars	1.242	0.870 to 1.772	0.232
Marginal bone loss
<1/3 root length	1	Reference
>1/3 root length	1.026	0.784 to 1.344	0.850

CI-confidence interval

Discussion

As in previously conducted surveys (, ), the present study sample was mainly composed of female subjects (68.2%). The median number of remaining teeth in the study sample was 26 (interquartile range 24-28), which is also comparable to the previous research (). Periapical disease was diagnosed in 5.4% of permanently restored non-endodontically treated teeth. Of the 597 examined participants, 54.3% had AP in one or more non-endodontically treated teeth. A high prevalence of AP may be attributable to the sampling method. This was a clinically-based study and the sample was drawn from the patients referred to the dental clinic, hence it is likely that the prevalence of periapical disease is higher than the prevalence expected in general population (). Yet, this study aimed to disclose the relation between the type of coronal restoration and periapical disease in non-endodontically treated teeth, and the method of sampling was not considered to have a crucial effect on the results.

The present survey focused on the quality of coronal restoration and periapical status at the time of the examination. The hypothesis was that there is no significant difference in the periapical status of non-endodontically treated teeth regarding the type or material of permanent coronal restoration. Indeed, the analysis of teeth with adequate quality of coronal restoration did not found any significant difference in the frequency of AP regarding the type of coronal restoration. Although some experimental studies have revealed some harmful effects of resin composite on pulpal cells, this does not seem to be the case when it is used for restorations in clinical conditions (, , ). Our study demonstrated that there are no differences in the prevalence of AP between teeth restored with amalgam and resin composite, irrelevant of their quality, which is consistent with the results of the previously cited study ().

When periapical status was analyzed according to the tooth type, only the group of premolars demonstrated a significant difference regarding the type of coronal restoration. AP was more frequent in premolars restored with crowns, than in those restored with amalgam or composite fillings. An interesting observation was that frontal and molar teeth restored with crowns of adequate quality had no periapical diseases. In contrast, almost 12% of premolars with adequate crowns had AP. Arguably, cumulative insults on dental pulp, such as caries, restorative treatments, periodontal disease and trauma may have diminished a long-term prognosis for pulpal vitality in these teeth (). They had possibly impaired the pulp ability to recover from further trauma that resulted from preparation to receive crowns. Furthermore, tooth preparation for full coverage crowns necessitates more extensive tooth structure removal than for other restoration types. The integrity and vitality of the pulp may also be impaired due to desiccation of dentin and development of frictional heat or exothermic reactions during setting of impression materials, acrylic resins and luting cements (). While certain studies found no variation in incidence of the pulp necrosis regarding the tooth type, Cheung at al. reported that majority of pulpal necroses developed in the maxillary anterior teeth that served as bridge abutments (, ).

The results regarding the differences in periapical status according to the type and quality of coronal restoration, as well as tooth type, were obtained using a bivariate analysis (chi-square test). Since more than one variable affects the course of disease, multivariate models are better approximation to reality, allowing for judgement of the relative importance of each predictor variable. There are variables that could have influenced the results but could not have been controlled due to the cross-sectional nature of the study, such as time of restoration placement, and pulpal and periapical status that preceded the restoration placement. Therefore, this study could only identify risk indicators for AP presence in restored teeth. A variable is designated as a risk indicator when its presence increases the individual’s risk of exhibiting periapical disease. However, risk indicators may, but need not, be directly associated with the disease (, ). Bearing in mind the previously mentioned limitations, logistic regression was used to analyze the simultaneous influence of predictor variables on AP presence. This analysis disclosed several possible risk indicators for AP presence. An increased risk for AP presence was found for premolars, crowned teeth and teeth with clinical or radiological diagnosis of recurrent caries. Compared to group of incisors and canines, premolars were 1.7 times more likely to have AP. Teeth restored with crowns were 3.4 times more likely to have AP than teeth restored with amalgam fillings. Interestingly, teeth with inadequate marginal integrity of restorations had a slightly increased, but not statistically significant risk for AP presence. In turn, the presence of recurrent caries demonstrated the highest association and increased the risk for radiologic diagnosis of AP 4.7 times. Contrary to the previously conducted studies, marginal bone loss was not associated with the presence of periapical disease (, ). While full coverage crowns seemed to have a positive effect on periapical health in endodontically treated teeth (), the present results indicated crowns as a risk indicator for poor periapical status in non-endodontically treated teeth. A possible explanation for this finding may be that the teeth that were restored with crowns were originally more compromised due to caries or fractures, than the teeth that were restored with direct restorations and therefore more prone to develop AP. In such a case, the AP would not be related to crowns but to the pre-existing compromised tooth structure.

A study by Hommez et al. demonstrated that the evaluation of coronal restoration quality based merely on radiographic images analysis did not provide adequate data; hence their study emphasized the importance of supplementation of radiographic information with clinical data (). In the present research, the assessment of the quality of the coronal restorations was based on both clinical and radiographic recordings.

This survey has several issues that might have led to underestimation of periapical pathosis. First, the cross-sectional design of the study precludes the follow-up of the periapical disease dynamics, while the frequency of AP depends on a period for the disease progress or regress. In the present study, PAI was used for AP diagnosis. A strict calibration procedure of observers and the use of visual references with verified histologic diagnoses enable higher reproducibility of the radiographic scores (, ). However, as in other studies of similar design, the present survey did not involve clinical criteria for AP diagnosis, and this might inevitably lead to underestimation of periapical pathosis (-, -). Since histological studies demonstrated that periapical inflammation may be associated with absence of radiographic signs (), it is important to use the most sensitive diagnostic tool to identify AP. Cone beam computed tomography (CBCT) has excellent accuracy in AP diagnostics when compared to conventional radiographs (). However, as it results in a higher radiation dose, CBCT use in endodontics should be limited to cases when conventional radiographic techniques do not provide sufficient information for confident diagnosis and treatment planning (, ). Due to lower radiation dose compared with a full-mouth survey, panoramic radiographs were chosen for the evaluation of periapical status in the present study. It has been shown that the prevalence of AP may be underestimated when compared to periapical radiography. However, when compared to cone beam computed tomography, both panoramic and periapical radiographs correctly identify AP in advanced stages ().

Several comprehensive studies regarding oral health that preceded our research utilized the panoramic radiographs as a radiographic technique of choice for periapical status evaluation (, , ). Panoramic radiographs allow the examination of all teeth in a single radiographic image under low radiation doses when compared to full mouth periapical radiography, and therefore are acceptable and frequently used for the evaluation of AP in epidemiological studies (, , -).

Although the data reported in the present study suggested that crowns may be a risk indicator for the presence of AP in non-endodontically treated teeth, the fact that the pulpal and periapical status at the time of restoration placement were unknown makes impossible to know if the AP was pre-existing or developed after the restoration was inserted. Therefore, to determine a causal-effective relationship between the type of coronal restoration and periapical disease, a prospective design of study should be applied, preferably by using a limited cone beam computed tomography for AP diagnosis.

Conclusion

The results of this research supported our hypothesis. If the quality of restorations is adequate, there is no difference in the periapical status of non-endodontically treated teeth regarding the type and material of coronal restoration.