Retained tooth root fragments, also known as retained root tips, are root fragments that remain in alveoli after loss of the more coronal portion of the tooth. Tooth root retention can result from various causes including dentoalveolar trauma, incomplete extraction, advanced caries, and end-stage tooth resorption.1–5 An RTRF can act as a nidus of inflammation and infection and may cause draining tracts, osteomyelitis, gingival inflammation, infection, and chronic pain.2,3,5–7 In an experimental model involving rats, RTRFs delayed healing at the tooth extraction site.8 In another study,9 31 of 226 (13.7%) dogs examined by a referral veterinary dentistry service and 31 of 123 (25.2%) dogs with clinically absent teeth had radiographic evidence of RTRFs, which indicates the condition is common in dogs.
The human dental literature indicates that the prevalence of clinical or radiographic evidence of inflammation associated with RTRFs is low.10,11 To our knowledge, studies are lacking regarding the prevalence of clinical and radiographic evidence of inflammation associated with RTRFs in dogs, and literature is sparse regarding clinical and radiographic sequelae associated with RTRFs in dogs. In 1 study,5 61 of 74 (82.4%) dogs that had undergone carnassial tooth extraction had RTRFs, of which 39 of 74 (52.7%) had radiolucency associated with the retained roots. That study5 was retrospective in nature, specifically evaluated dogs and sites with a known history of tooth extraction, and did not assess clinical findings. Because veterinary patients are unable to describe what they are feeling, clinical signs associated with pain in the oral cavity are frequently not reported.12 Consequently, sequelae associated with tooth root retention may be underappreciated in dogs.
When extraction of a tooth is necessary, clinicians should always strive to remove the entire tooth, and if iatrogenic root fracture occurs during extraction, the root fragment should be retrieved and removed.7 However, there are circumstances in which the risk to the patient owing to prolonged anesthesia or the potential of damage to adjacent structures might outweigh the benefit of root retrieval. In those situations, it is necessary for clinicians to decide whether to attempt root extraction or allow the root fragment to remain in situ with frequent monitoring.7 Indications for extraction include an RTRF length > 4 mm,7 retention of the root superficially within the alveolus,7,13 presence of clinical or radiographic signs of inflammation,3,7,13,14 and mobilization of the fragment.11 However, an association between RTRF size and the presence of clinical or radiographic evidence of inflammation has not been established in dogs. Moreover, although superficially located RTRFs are associated with an increased risk for clinical or radiographic evidence of inflammation in human patients,10,15 it is unknown whether the same is true for dogs.
Retrieval of RTRFs can be technically challenging, and complications associated with the procedure are potentially serious and can be exacerbated by poor surgical technique. The procedure can result in penetration of adjacent spaces, such as the nasal cavity, maxillary recess, and infraorbital and mandibular canals, and may lead to displacement of the RTRF into those spaces or damage to associated neurovascular structures.13,16,17 Penetrating injuries to the orbit or cranium secondary to the use of inappropriate tooth extraction techniques in the caudal portion of the maxilla have been reported and are potentially catastrophic.18,19 Therefore, validation of criteria used to determine whether RTRFs should be extracted or allowed to remain in situ is necessary. The purpose of the study reported here was to assess the frequency of clinical and radiographic evidence of inflammation associated with RTRFs in dogs and to determine whether clinical or radiographic evidence of inflammation was affected by RTRF length and position within the alveolar bone.
Materials and Methods
Animals
The study had a cross-sectional design. All study data were obtained as part of the clinical and diagnostic workup of client-owned dogs with the owners’ consent; therefore, the study was exempt from institutional animal care and use committee approval.
Dogs > 1 year old that underwent full-mouth radiographic examination for the diagnosis of oral disease at Perth Pet Dentistry, Perth, Australia, between September 2015 and June 2016 and at the Cornell University Hospital for Animals, Ithaca, NY, between July 2016 and December 2017 were considered for study enrollment. Each dog enrolled in the study had to have evidence of an RTRF, which was defined as any remaining part of a tooth apical to the cementoenamel junction after the crown of the tooth was removed or lost. Each dog included in the study also had to have a complete medical record and documentation of a comprehensive oral examination, which included periodontal probing and charting. Dogs with tooth root fragments with portions of the crown attached and dogs with root fragments with radiographic evidence of extensive root resorption such that the fragment could not be easily identified were excluded from the study. Dogs referred for RTRF retrieval that had undergone crown loss or a failed extraction attempt < 1 month prior to examination by the dental service were also excluded from the study.
Data collection
All oral radiographic images were obtained with digital dental radiography units and phosphor plate systemsa,b by use of standardized intraoral techniques as described.20 Diagnosis of an RTRF was made on the basis of the clinical absence of a crown and the presence of a radiopaque structure in the region of a missing tooth with the characteristic shape and radiopacity of a tooth root.14
For each dog enrolled in the study, the medical record was reviewed and signalment data were extracted. Because of the unreliability of owner recollection and inconsistency in recording of patient history, data regarding attempted tooth extraction or apparent loss of tooth crowns were not extracted. The number and location of RTRFs were recorded for each dog. Recorded information included the side of the oral cavity where the RTRF was located and whether the fragment was located in the maxilla or mandible at one of the following regions: incisor, canine, rostral premolar (excluding the maxillary fourth premolar region), or carnassial-molar region. When possible, RTRFs were classified as deciduous or permanent. Deciduous fragments were identified on the basis of their appearance and location in conjunction with the presence of a corresponding permanent tooth at the same location. It was not always possible to identify the specific tooth or root involved for every RTRF identified. Therefore, the location and size of each RTRF were independently recorded regardless of whether multiple fragments were parts of the same tooth.
For each RTRF identified, 1 investigator (KKN) extracted information regarding the clinical appearance of the fragment from the medical record and evaluated and obtained measurements from the corresponding radiographic images. The position of the RTRF was classified as protruding (coronal end of the fragment superficial to the alveolar bone margin), level (coronal end of the fragment even with the alveolar bone margin), or buried (coronal end of the fragment deep to the alveolar bone margin) on the basis of review of radiographic images. The apico-coronal length of each RTRF was measured on digital radiographic images by use of softwarec,d provided with the radiography systems. The presence or absence of radiographic evidence of inflammation was noted. For the purpose of this study, radiographic findings consistent with inflammation included radiolucency associated with the RTRF, an abnormally wide pulp cavity or periodontal ligament, vertical alveolar bone loss, and inflammatory root resorption. Intraoral and extraoral clinical findings were also recorded. Intraoral findings included the appearance of the gingiva overlying the RTRF and the presence or absence of an intraoral draining tract. The gingival appearance was categorized as normal, inflamed without a visible root fragment or draining tract, inflamed with the presence of a draining tract but no root fragment visible, or inflamed with the root fragment visible. The presence of an intraoral draining tract in an area other than directly over the RTRF, such as the mucogingival junction, was also recorded. Extraoral clinical findings recorded included the presence or absence of palpable extraoral swelling or an extraoral draining tract.
Statistical analysis
Descriptive data were generated. The frequency of occurrence was reported for categorical variables. The median, range, and IQR were reported for continuous variables. Generalized linear mixed models with a binomial distribution and logit link were used to identify factors associated with the presence of clinical and radiographic evidence of inflammation or both (ie, evidence of inflammation). Each model included a random effect for dog to account for nonindependence of RTRF measurements. Three sets of fixed effects were considered (RTRF length, RTRF length and position, and RTRF length, position, and the interaction between RTRF length and position). The goodness of fit was compared among competing models (ie, sets of fixed effects) by means of the likelihood ratio test. Results were reported as the ORs and corresponding 95% CIs. For models that included RTRF position as a fixed effect, the Tukey method was used to correct for type I error inflation when post hoc pairwise comparisons were necessary. Values of P ≤ 0.05 were considered significant.
An ROC curve was plotted to assess the usefulness of various cutoff values of RTRF length as predictors for the presence of clinical and radiographic evidence of inflammation. The AUC was calculated by use of the DeLong method. The ability of a given RTRF length cutoff to predict the presence of clinical or radiographic evidence of inflammation was considered excellent if the AUC was ≥ 0.9 to 1, good if the AUC was ≥ 0.8 to < 0.9, fair if the AUC was ≥ 0.7 to < 0.8, poor if the AUC was ≥ 0.6 to < 0.7, and unacceptable if the AUC was ≥ 0.5 to < 0.6. All statistical analyses were performed by use of an open-source statistics program.e
Results
Dogs
During the study period, full-mouth radiographic evaluation was performed for 383 dogs, of which 85 (22.2%) had at least 1 RTRF. The prevalence of RTRFs was similar between the Australian (41/179 [22.9%]) and US (44/204 [21.6%]) study institutions. Sixty-six dogs with a total of 148 RTRFs met the inclusion criteria and were enrolled in the study. Three dogs were excluded because of crown loss or a failed extraction attempt < 1 month prior to examination at the study institution, and 16 dogs were excluded owing to insufficient detail in the medical record regarding the clinical appearance of the RTRF site.
For the 66 dogs included in the study, the median age was 10.5 years (range, 1 to 17 years; IQR, 8 to 12.8 years) and body weight was 9.4 kg (20.7 lb; range, 3.0 to 48.6 kg [6.6 to 106.9 lb]; IQR, 6.1 to 16.8 kg [13.4 to 37.0 lb]). The study population consisted of 28 (42%) neutered males, 3 (5%) sexually intact males, 33 (50%) spayed females, and 2 (3%) sexually intact females.
RTRFs
The median number of RTRFs identified per dog was 2 (range, 1 to 13; IQR, 1 to 2). One hundred forty-eight RTRFs were identified, of which 86 (58%) were in the maxilla and 62 (42%) were in the mandible. An equal number (n = 74) of RTRFs were located on the left and right sides of the oral cavity. Retained tooth root fragments were most frequently located in the rostral premolar region (n = 62 [41.9%]), followed by the carnassial-molar region (42 [28.4%]) and incisor region (38 [25.7%]). Only 6 (4%) RTRFs were located in the canine region, and 5 of those 6 fragments were classified as being from deciduous teeth. Root fragments from deciduous teeth were not definitively identified in any other region.
The median length of RTRFs was 5.7 mm (range, 1.1 to 37.6 mm; IQR, 4.1 to 8.3 mm). The longest RTRF (37.6 mm) was from a permanent tooth located in the canine region. The next longest RTRF (18.2 mm) was located in the incisor region. All statistical analyses were performed with and without the longest RTRF included in the data set. Differences in results were minimal with and without the longest RTRF included in the data set; therefore, all results reported henceforth are for analyses that included that fragment.
The position of the fragment relative to the alveolar bone margin was classified as buried for 64 (43.2%; Figure 1) of the 148 RTRFs, level for 31 (20.9%; Figure 2), and protruding for 53 (35.8%; Figure 3). Radiographic evidence of inflammation was identified for 59 (39.9%) RTRFs.
Representative intraoral radiographic images obtained from dogs with RTRFs that were classified as buried (ie, the coronal end of the fragment was located deep to the alveolar bone margin) with and without evidence of inflammation. A—Image of a buried distal root fragment of a right mandibular premolar tooth (arrow) without evidence of inflammation. B—Image of buried mesial (arrow) and distal (arrowhead) root fragments of a right mandibular fourth premolar tooth. Notice the abnormally wide periodontal ligament and radiolucency centered at the coronal end of the mesial root fragment, which were characteristic of periodontitis; no evidence of inflammation is associated with the distal root fragment. C—Image of buried mesiobuccal (arrow) and distal (arrowhead) root fragments of a left maxillary fourth premolar tooth. Notice the large radiolucency surrounding both fragments, as well as the irregular appearance of the distal root fragment, which was indicative of inflammatory root resorption.
Citation: Journal of the American Veterinary Medical Association 256, 6; 10.2460/javma.256.6.687
Representative intraoral radiographic images obtained from dogs with RTRFs that were classified as level (ie, the coronal end of the fragment was located even with the alveolar bone margin) with and without evidence of inflammation. A—Image of a level mesially located tooth root fragment (arrow) located in the left mandibular premolar region that has no evidence of inflammation. Notice that a more distally located root fragment (arrowhead) is protruding from the alveolar margin and has evidence of vertical bone loss. B—Image of a level root fragment of a left maxillary first premolar tooth (large arrow) with periapical radiolucency suggestive of inflammation. Also present are protruding mesial (arrowhead) and level distal (small arrow) root fragments of a left maxillary second premolar tooth; there is periapical radiolucency associated with the mesial fragment, but no evidence of inflammation associated with the distal fragment. C— Image of a level distal root fragment (arrow) of a left mandibular fourth premolar tooth with multiple fragments and vertical bone loss.
Citation: Journal of the American Veterinary Medical Association 256, 6; 10.2460/javma.256.6.687
Representative intraoral radiographic images obtained from dogs with RTRFs that were classified as protruding (ie, the coronal end of the fragment was located superficial to the alveolar bone margin) with and without evidence of inflammation. A—Image of protruding root fragments of the right first (arrowhead), second (large arrow), and third (small arrow) incisor teeth. The first incisor root fragment has an abnormally wide pulp cavity and end-stage vertical bone loss, and there is a larger periapical radiolucency associated with the third incisor root fragment. The second incisor root fragment has no evidence of inflammation. B—Image of a protruding distal root fragment of a right maxillary second premolar tooth with a periapical radiolucency (arrow). C—Image of a protruding mesial root fragment of a right mandibular third premolar tooth (arrow) with vertical bone loss. A level distal root fragment of the right mandibular third premolar tooth (arrowhead) is present but difficult to discern owing to the presence of ankylosis.
Citation: Journal of the American Veterinary Medical Association 256, 6; 10.2460/javma.256.6.687
Sixty-eight of the 148 (45.9%) RTRFs had clinical evidence of inflammation. The most common inflammation-associated clinical finding was the presence of a draining tract in the gingiva overlying a root fragment (n = 30 [44.1%]), followed by a visible root fragment (22 [32.4%]) and inflamed gingiva overlying a root fragment (16 [23.5%]; Figure 4). Intraoral draining tracts were not identified in any location other than directly over an RTRF. Extraoral swelling was associated with 4 of the 148 (2.7%) RTRFs, but no extraoral draining tracts were identified. The majority (46/68 [67.6%]) of RTRFs with clinical evidence of inflammation also had radiographic evidence of inflammation. Only 13 of 80 (16.3%) RTRFs without clinical evidence of inflammation had radiographic evidence of inflammation. Collectively, 81 of 148 (54.7%) RTRFs had clinical or radiographic evidence of inflammation or both (ie, evidence of inflammation). Evidence of inflammation was identified for 30 of 64 (46.9%) buried, 11 of 31 (35.5%) level, and 40 of 53 (75.5%) protruding root fragments.
Representative photographs of the oral cavities of dogs with RTRFs with and without clinical evidence of inflammation. A— Photograph of the right maxillary premolar region of the dog in Figure 3B. Notice that the gingiva overlying the protruding distal root fragment of the right maxillary second premolar tooth (arrow) appears normal with no evidence of inflammation. B— Photograph of the rostral portion of the maxillary dental arch of the dog in Figure 3A. The gingiva overlying the protruding root fragment of the right maxillary second incisor appears normal (large arrow). The protruding root fragment of the right maxillary first incisor tooth is visible (arrowhead), and there is a draining tract overlying the protruding root fragment of the right maxillary third incisor tooth (small arrow). C—Photograph of the buccal aspect of the left mandibular premolar region of the dog in Figure 2A. The tip of the dental probe indicates the location of a 3-mm draining tract overlying the most distal protruding RTRF. D—Photograph of the buccal aspect of the left side of the oral cavity of the dog in Figure 1C. There are 2 draining tracts with white purulent discharge (arrows) overlying the mesiobuccal and distal root fragments of the left maxillary fourth premolar tooth. E—Photograph of the buccal aspect of the left maxillary premolar region of the dog in Figure 2B. There is a small area of inflamed gingiva overlying the level root fragment of the left maxillary first premolar tooth (large arrow). The protruding mesial root fragment (arrowhead) of the left maxillary second premolar tooth is visible, whereas the gingiva overlying the level distal root fragment (small arrow) appears clinically normal. F—Photograph of the right mandibular premolar region of the dog in Figure 3C. The protruding mesial root fragment of the right mandibular third premolar tooth (arrow) is visible, and the gingiva overlying the level distal root fragment (arrowhead) is inflamed with a cobblestone appearance.
Citation: Journal of the American Veterinary Medical Association 256, 6; 10.2460/javma.256.6.687
The goodness of fit for the generalized linear mixed model that included RTRF length and position was significantly (P < 0.001) better than that for the model that included only RTRF length. The addition of a fixed effect for the interaction between RTRF length and position did not significantly (P = 0.826) improve the goodness of fit for the model. Those findings suggested that both RTRF length and position were significantly associated with the presence of inflammation, but the effect of RTRF length on the presence of inflammation did not differ on the basis of RTRF position (buried, level, or protruding) relative to the alveolar bone margin. Therefore, results were reported only for the generalized linear mixed model that included fixed effects for RTRF length and position (the model with 2 fixed effects). The odds of inflammation for protruding RTRFs were 2.98 (95% CI, 1.02 to 8.72; P = 0.046) and 7.58 (95% CI, 1.98 to 29.08; P = 0.001) times those for buried and level RTRFs, respectively. The odds of inflammation did not differ significantly (P = 0.179) between buried and level RTRFs (OR, 2.55; 95% CI, 0.74 to 8.78). For every 1-mm increase in RTRF length, the odds of inflammation increased by 17% (OR, 1.17; 95% CI, 1.04 to 1.34; P = 0.009).
The sensitivity and specificity for various RTRF length cutoffs for prediction of the presence of inflammation were summarized (Table 1) and used to plot an ROC curve (Figure 5). The AUC was 0.61 (95% CI, 0.52 to 0.70; P = 0.024), which indicated that RTRF length was a poor predictor of inflammation.
Sensitivity and specificity for various tooth root fragment length cutoffs when used to predict the presence of clinical or radiographic evidence of inflammation as determined by analysis of data for 148 RTRFs in 66 dogs.
| Tooth root fragment length cutoff (mm) | Sensitivity (%) | Specificity (%) |
|---|---|---|
| 1 | 100 | 0 |
| 2 | 95 | 3 |
| 3 | 86 | 12 |
| 4 | 78 | 22 |
| 5 | 67 | 43 |
| 6 | 56 | 61 |
| 7 | 48 | 67 |
| 8 | 41 | 81 |
| 9 | 30 | 88 |
| 10 | 21 | 94 |
Receiver operating characteristic curve that depicts the discriminatory ability of RTRF length for predicting the presence of clinical or radiographic evidence of inflammation. Notice that the curve is located close to the identity (diagonal) line, which indicates that the ability of root fragment length to predict the presence of clinical or radiographic evidence of inflammation is only slightly better than chance. The ROC curve was generated from data for 148 RTRFs in 66 dogs, and the AUC was 0.61 (95% CI, 0.52 to 0.70).
Citation: Journal of the American Veterinary Medical Association 256, 6; 10.2460/javma.256.6.687
Discussion
In the present study, 85 of 383 (22.2%) dogs that underwent a full-mouth radiographic evaluation at 1 of 2 referral dental practices during the observation period had at least 1 RTRF. That prevalence was greater than the prevalence of dogs with RTRFs (31/226 [13.7%]) reported in another study9 and underscored the clinical importance of RTRFs. The present study population was skewed toward older (median age, 10.5 years) and fairly lightweight (median body weight, 9.4 kg) dogs. Factors that may increase the risk of tooth root fracture and retention include the fairly high prevalence of periodontitis in small and senior dogs,21 the small tooth size of small dogs, and increased brittleness of dentin with increasing age.22 However, it is possible that the demographic characteristics reported for the dogs of the present study were specific to that population and did not reflect those of the general population of dogs with RTRFs. For the dogs of the present study, RTRFs were more frequently identified in the maxilla than in the mandible and the rostral premolar region, compared with other locations. Those findings may reflect the large number of teeth in those locations or indicate that the teeth at those locations are more susceptible to trauma than are teeth at other regions. Unfortunately, no conclusions can be drawn regarding an association between tooth location or type and the risk for development of RTRFs on the basis of the results of the present study. Further research is necessary to adequately evaluate the effects of dog age and body weight and tooth type and location on the development of RTRFs.
The majority (81/148 [54.7%]) of RTRFs evaluated in the present study had clinical or radiographic evidence of inflammation, with 46 (31.1%) fragments having both clinical and radiographic evidence of inflammation (evidence of inflammation). The high prevalence of RTRFs with evidence of inflammation in the present study supported current recommendations for complete extraction of teeth whenever possible. In a retrospective study5 of carnassial tooth extraction sites in dogs and cats, 39 of 74 (52.7%) dogs and 27 of 42 (64.3%) cats with RTRFs had radiographic evidence of periapical abnormalities. The reported prevalence of human patients with RTRFs and clinical or radiographic evidence of inflammation (22.5% and 27.2%)10,11 is much lower than that reported in veterinary patients. The higher prevalence of RTRFs with clinical or radiographic evidence of inflammation in veterinary patients relative to human patients is likely a reflection of many factors, including interspecies differences in the etiologic factors associated with the development of RTRFs, differences in study methodology, the inability of veterinary patients to describe or report clinical signs and seek intervention, and differences in the tooth extraction skills of veterinarians and human dentists. That last factor may be related to the paucity of instruction in dentistry and oral surgery received by veterinarians.23
In human dentistry, there is a positive association between the proximity of the RTRF to the gingival surface and likelihood of clinical or radiographic evidence of inflammation.10,15 In 1 human study,10 radiographic abnormalities were identified for 199 of 334 (59.6%) root fragments exposed to saliva versus only 75 of 1,666 (4.5%) root fragments unexposed to saliva. Similarly, in another human study,15 radiolucencies were associated with 438 of 540 (81.1%) root fragments exposed to the oral cavity, compared with only 86 of 419 (20.5%) root fragments not exposed to the oral cavity.
Surgical removal of superficially located RTRFs, particularly those resulting from iatrogenically induced root fractures, is currently recommended in both human13 and veterinary7 dentistry. Results of the present study supported that recommendation because RTRFs that protruded from the alveolar bone margin were more likely to have evidence of inflammation than were RTRFs buried (ie, deep) or level with the alveolar bone margin. Protrusion of an RTRF beyond the alveolar bone margin likely predisposes the overlying gingiva to trauma from mastication, which may result in previously intact gingiva fenestrating, thus resulting in exposure of the root fragment to the oral microbiome. Subsequent colonization of the exposed pulp or root surface by oral microbes may result in periapical abnormalities and periodontitis, respectively. Although the odds of inflammation did not differ significantly between buried and level RTRFs in the present study, root fragments located level with the alveolar bone margin are more likely to become exposed to the oral cavity as a result of gradual alveolar bone resorption.11,24 Therefore, clinicians should consider removal of RTRFs whenever possible, especially those that are level with or protruding from the alveolar bone margin.
Review of both the veterinary and human medical literature revealed that only 1 human study15 has evaluated the relationship between the size of an RTRF and the presence of radiographic abnormalities. In that study,15 the frequency of radiographic abnormalities did not differ significantly between RTRFs that were smaller and larger than a third of the original root length. In the present study, the odds of inflammation increased by 17% for each 1-mm increase in RTRF length, and that effect was not dependent on the position of the fragment relative to the alveolar bone margin. That finding was somewhat surprising because the human dental literature10,15 suggests that exposure of RTRFs to the oral cavity is the primary driver for inflammatory sequelae. Therefore, we anticipated that RTRFs that protruded past the alveolar bone margin were more likely to have evidence of inflammation than were fragments that were buried or level with the alveolar bone margin, regardless of fragment length. It is possible that increasing root fragment size increases both the periodontal and endodontic surface area for the accumulation of bacterial biofilm.
It is currently recommended that RTRFs > 4 mm in size be removed.7 In the present study, ROC analysis revealed that there was no specific root fragment length that could accurately distinguish between fragments with and without evidence of inflammation with both high sensitivity and specificity. Nevertheless, sensitivity increased and specificity decreased as RTRF length decreased. Thus, although the results of the present study supported the recommendation for removal of larger RTRFs, the definition of larger (ie, the root fragment size criterion used to definitively decide that extraction is the best course of action) appeared to be arbitrary. Therefore, we recommend complete removal of RTRFs whenever possible, and if complete removal of the fragment cannot be achieved, we recommend that as much of the fragment be removed as feasible without compromising the adjacent tissues or health of the patient.
Retained tooth root fragments that are small, are deeply embedded, have no evidence of inflammation, and have not been mobilized by failed extraction attempts can be left in situ if extraction is likely to compromise surrounding tissues or the patient is unstable. In such cases, clients should be informed of the RTRF, an intraoral radiograph of the fragment should be obtained to document its location and position for future reference, and the patient and RTRF should be monitored clinically and radiographically periodically.7,16 However, it is important to note that, in the present study, only 4 of 81 (4.9%) RTRFs with evidence of inflammation were associated with detectable extraoral abnormalities and 13 of the 80 (16.3%) RTRFs that did not have clinical evidence of inflammation had radiographic evidence of inflammation. Therefore, for dogs in which RTRFs are intentionally left in situ, periodic clinical and radiographic oral evaluations with the patient anesthetized should be prioritized over owner-based monitoring and oral examination while the patient is conscious.
The present study was not without limitations. The bisecting-angle technique was used to obtain all intraoral radiographs except for images of the caudal aspect of the mandible. That technique might have caused distortion of the RTRFs on the radiographic images and resulted in misclassification of some protruding root fragments as buried or level because of superimposition of the alveolar bone over the root fragment or inaccurate measurement of some root fragments. Standardized techniques were used to obtain all intraoral radiographs and minimize root fragment distortion. Cone-beam CT would have been a more accurate method than intraoral radiography for identification and measurement of RTRF position and length. However, that diagnostic imaging modality was considered impractical for the purpose of this study. Intraoral radiographs can be easily obtained and are inexpensive. Moreover, the intraoral radiographic techniques used to obtain the images for this study reflect those commonly used in veterinary dental practices and thereby increase the applicability of our conclusions in clinical settings.
In the present study, clinical and full-mouth radiographic findings were evaluated for all RTRFs identified, and it was not possible to determine whether the presence of periodontitis or periapical abnormalities contributed to or were sequelae of root fragment retention. Thus, the findings of the present study should be applied with caution to patients that have intentional root fragment retention. In a study10 of human patients, histologic examination of 60 suspected RTRFs revealed that some specimens were actually sclerotic bone or odontomas. It is possible that some of the RTRFs evaluated in the present study were instead islands of dense bone or sclerosis, and histologic evaluation would have been useful to confirm RTRF diagnosis as well as to assess pulp vitality.
The majority (81/148 [54.7%]) of RTRFs identified in the dogs of the present study were associated with clinical or radiographic evidence of inflammation that was sufficient to constitute an indication for extraction. Extraoral clinical abnormalities were rare, and a detailed intraoral examination in conjunction with the acquisition of full-mouth intraoral radiographs was necessary for accurate diagnosis of RTRFs. Increasing RTRF size and protrusion of the fragment beyond the alveolar bone margin were associated with increased odds of inflammation. Thus, results of the present study supported the current recommendation for extraction of RTRFs whenever feasible.
Acknowledgments
The authors declare that there were no conflicts of interest.
The authors thank Stephen Parry for assistance with data analyses.
ABBREVIATIONS
| AUC | Area under the receiver operating characteristic curve |
| CI | Confidence interval |
| IQR | Interquartile (25th to 75th percentile) range |
| ROC | Receiver operating characteristic |
| RTRF | Retained tooth root fragment |
Footnotes
CS 7600, Carestream Dental, Atlanta, Ga.
CR7 Vet, Dürr Medical, Bietigheim-Bissingen, Germany.
Dental Imaging Software, Carestream Dental, Atlanta, Ga.
Vetexam, Dürr Medical, Bietigheim-Bissingen, Germany.
R, version 3.4.0, R Foundation for Statistical Computing, Vienna, Austria.
References
1. DuPont G. Crown amputation with intentional root retention for cases of advanced feline resorptive lesions—a clinical study. J Vet Dent 1995;12:9–13.
2. Galante KM, Beard D. What Is Your Diagnosis? Retained root fragment of the left mandibular canine tooth (tooth No. 304). J Am Vet Med Assoc 2004;225:1037–1038.
3. Holmstrom SE, Frost P, Eisner ER. Exodontics. In: Veterinary dental techniques for the small animal practitioner. 3rd ed. Philadelphia: Saunders, 2004;291–338.
4. Woodward TM. Extraction of fractured tooth roots. J Vet Dent 2006;23:126–129.
5. Moore JI, Niemiec B. Evaluation of extraction sites for evidence of retained tooth roots and periapical pathology. J Am Anim Hosp Assoc 2014;50:77–82.
6. Kapatkin AS, Marretta SM, Patnaik AK, et al. Mandibular swellings in cats: perspective study of 24 cats. J Am Anim Hosp Assoc 1991;27:575–580.
7. Lommer MJ. Principles of exodontics. In: Verstraete FJM, Lommer MJ, eds. Oral and maxillofacial surgery in dogs and cats. St Louis: Saunders, 2012;97–114.
8. Glickman I, Pruzanskly S, Ostrach M. The healing of extraction wounds in the presence of root remnants and bone fragments. Am J Orthod 1947;33:263–283.
9. Verstraete FJM, Kass PH, Terpak CH. Diagnostic value of full-mouth radiography in dogs. Am J Vet Res 1998;59:686–691.
10. Helsham RW. Some observations on the subject of roots of teeth retained in the jaws as a result of incomplete exodontia. Aust Dent J 1960;5:70–77.
11. Herd JR. The retained tooth root. Aust Dent J 1973;18:125–131.
12. Richey M. Anesthesia and pain management in dental and oral procedures. In: Holmstrom SE, Frost P, Eisner ER, eds. Veterinary dental techniques for the small animal practitioner. 3rd ed. Philadelphia: Saunders, 2004;601–624.
13. Hupp JR. Principles of more complex exodontia. In: Hupp JR, Ellis E, Tucker MR, eds. Contemporary oral and maxillofacial surgery. 6th ed. St Louis: Mosby, 2014;119–142.
14. Dupont GA, Debowes LJ. Miscellaneous conditions. In: Atlas of dental radiography in dogs and cats. Philadelphia: Saunders, 2009;221–228.
15. Dachi SF, Howell FV. A survey of 3,874 routine full-mouth radiographs I. A study of retained roots and teeth. Oral Surg Oral Med Oral Pathol 1961;14:916–924.
16. Lommer MJ. Complications of extractions. In: Verstraete FJM, Lommer MJ, eds. Oral and maxillofacial surgery in dogs and cats. St Louis: Saunders, 2012;153–159.
17. Taylor TN, Smith MM, Snyder L. Nasal displacement of a tooth root in a dog. J Vet Dent 2004;21:222–225.
18. Duke FD, Snyder CJ, Bentley E, et al. Ocular trauma originating from within the oral cavity: clinical relevance and histologic findings in 10 cases (2003–2013). J Vet Dent 2014;31:245–248.
19. Troxel M. Iatrogenic traumatic brain injury during tooth extraction. J Am Anim Hosp Assoc 2015;51:114–118.
20. Tsugawa AJ, Verstraete FJM. How to obtain and interpret periodontal radiographs in dogs. Clin Tech Small Anim Pract 2000;15:204–210.
21. Harvey CE, Shofer FS, Laster L. Association of age and body weight with periodontal disease in North American dogs. J Vet Dent 1994;11:94–105.
22. Arola D, Reprogel RK. Effects of aging on the mechanical behaviour of human dentin. Biomaterials 2005;26:4051–4061.
23. Anderson JG, Goldstein G, Boudreaux K, et al. The state of veterinary dental education in North America, Canada, and the Caribbean: a descriptive study. J Vet Med Educ 2017;44:358–363.
24. von Wowern N, Winther S. Submergence of roots for alveolar ridge preservation. A failure (4-year follow-up study). Int J Oral Surg 1981;10:247–250.
