In some situations, extraction of maxillary or mandibular second, third, and fourth premolar or first, second, and third molar teeth (ie, cheek teeth) is indicated to manage dental conditions in equids such as periapical infection, dental fracture, pulp exposure, tooth displacement or malformation, presence of supernumerary teeth, odontogenic or paranasal neoplasia, or problematic diastemata causing periodontal disease.1–5 Four main approaches have been described to extract equine cheek teeth, including intraoral extraction,3 repulsion,6 lateral buccotomy,7 and MIB, with or without transbuccal screw placement.5 Dixon et al4 reported success rates (ie, complete removal of all the dental material) of up to 90% for intraoral extraction of 111 teeth from 100 horses (mean age, 8 years) and a success rate of 83% when a fracture was present in the tooth to be extracted. In a study8 performed to evaluate endoscopically guided extraction of 31 fractured cheek teeth from 30 horses (median age, 11.5 years), 27 (87%) affected teeth were successfully extracted. In recent investigations, modifications to the original intraoral extraction technique with partial coronectomy9 and partial reconstruction of fractured teeth by use of PMMA10 for stabilization prior to attempted removal yielded successful outcomes for 193 of 194 (99%) and 16 of 21 (76%) teeth, respectively. Postoperative complications are less common after intraoral extraction, compared with external transcutaneous approaches, because of less trauma to the alveolus and supporting bony structures.1,2,7
Radiography has moderate to high specificity (70% to 90%) but poor sensitivity (52% to 69%) for the assessment of dental disorders.11–13 By providing cross-sectional data with greater contrast resolution (ie, distinction between different intensities), CT facilitates identification of subtle dental changes within the periodontium, pulp, apices of the roots, and clinical crown.14–20 Computed tomography has been identified as highly sensitive and specific for recognition of periapical infection of the cheek teeth in horses.21,22 Results of a study23 that compared radiography and CT revealed that the latter had a sensitivity of 100% and a specificity of 96.7% for diagnosis of dental disease. The authors are not aware of previous studies investigating the influence of preoperative CT and clinical data on the success of intraoral dental extraction in horses and ponies. Such data would be very valuable for veterinarians and owners to make decisions regarding the choice of surgical approaches. The objectives of the study reported here were to describe clinical and CT findings for horses and ponies that underwent intraoral extraction of a cheek tooth under standing sedation and to assess potential associations between these features and outcome of the procedure. We hypothesized that preexisting abnormalities involving the apical and periapical structures identified on CT, such as presence of hypercementosis or severe alveolar bone sclerosis, and evidence of structural failures of the tooth, including fractures or severe infundibular caries, would be negatively associated with successful intraoral extraction.
Materials and Methods
Case selection criteria
The study was designed as a retrospective cohort study. Electronic and hard copy medical records of the Royal Veterinary College Equine Referral Hospital were searched to identify horses and ponies that underwent CT of the head and subsequent intraoral extraction of a cheek tooth with standing sedation between January 1, 2010, and December 31, 2015. In cases where 2 adjacent teeth were removed in a single visit, only the first extracted tooth was included in the study to eliminate any effect on extraction success of altered structural integrity and periodontal attachments of the second extracted tooth. In cases where 2 adjacent teeth were removed on different occasions, the second tooth was considered for inclusion in the study only if the time elapsed between visits was > 2 months and no complications from the first extraction had been confirmed; the 2-month interval was chosen as the expected time for complete healing of the alveolar socket following a dental extraction. Records were excluded if CT results were deemed nondiagnostic.
Medical records review
Keywords for the electronic medical records search included tooth, extraction, exodontia, and CT. Patient signalment (age, breed, and sex), the nature and duration of clinical signs, the main reason for the evaluation, identification of the tooth extracted, and outcome following intraoral extraction were recorded. For the horses or ponies that had more than 1 extraction during separate visits, the history and signalment information were recorded for each visit.
Intraoral extraction was considered successful when the entire target tooth was extracted without extraoral approaches; if the tooth fractured during the procedure, intraoral extraction success was determined by examination and reconstruction of extracted dental material, endoscopic examination of the alveolus, or probing (digitally or with an instrument), or with a combination of these methods. The main reason for evaluation was as reported in the record or the predominant clinical finding during the initial examination. When a dental fracture was seen during oral examination prior to the CT evaluation, this information was recorded.
CT scan acquisition and examination
The hospital protocol for CT scanning of standing, sedated horses and ponies was as previously described24 and consisted of premedication with acepromazine (0.02 mg/kg [0.009 mg/lb], IV), followed 20 minutes later with detomidine hydrochloride (0.01 mg/kg [0.005 mg/lb]) and butorphanol tartrate (0.01 mg/kg) IV. The CT scans were obtained with a 16-slice multidetector CT scanner.a Scan settings were as follows: exposure, 120 kVp and 200 mA; slice thickness, 1.25 mm; slice gap, 1.25 mm; pitch, 0.625; tube rotation time, 0.8 seconds; and field of view, 50 cm. Images were reconstructed with standard and bone algorithms, and images were reviewed with viewing softwareb on dedicated screens.
For each case included in the study, the tooth removed by means of intraoral extraction was identified from the medical records, and the corresponding CT scan was retrospectively evaluated and graded through the consensus interpretation of a large animal radiologist (JJD) and an equine surgeon (THW) with board certifications in their respective fields. The reviewers were blinded to the diagnosis and the surgical outcome reported in the medical records at the time of the CT scan evaluations. Multiplanar reconstructions were created, and adjustments to window width and window level were applied on a case-by-case basis to optimize diagnostic utility of the CT data set, replicating clinical practices.
Attention was focused on the CT appearance of the periodontium, the affected cheek tooth, and the ipsilateral sinus compartments (Figure 1) Periapical sclerosis was identified by increased attenuation or thickness (or both) of the alveolar bone surrounding a tooth root and was subjectively graded as absent (normal), mild, moderate, or severe for each root. A nondetectable lamina dura was recorded if a continuous bone-attenuating structure following the contour of a root or the unerupted portion of the tooth could not be recognized or if a lamina dura defect was clearly present. When the lamina dura was intact, the periodontal space was assessed at the widest point and described as normal or as mildly, moderately, or severely widened. Gas in the periodontal space or in a pulp chamber was recorded when present. Root clubbing, commonly associated with apical hypercementosis, was identified when a rounded root with increased attenuation, compared with that of the contralateral root, was observed (if the contralateral root was considered normal by CT assessment).
Representative transverse CT image obtained at the level of the right and left maxillary first molar teeth in an adult horse (14 years of age) in a retrospective study to describe clinical and CT findings for 74 horses and 7 ponies that underwent intraoral extraction of a cheek tooth with standing sedation and to assess potential associations between these features and procedure outcome (successful vs unsuccessful intraoral extraction). There is evidence of grade 3 and grade 2 infundibular caries in the rostral infundibula of the right and left maxillary first molar teeth, respectively (arrowheads), with moderate periapical sclerosis and widening of the periodontal spaces (arrows). The left rostral and caudal maxillary sinuses and ventral conchal sinus are filled with fluid, indicative of sinusitis (asterisks). L = Left. R = Right.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
Representative transverse CT image obtained at the level of the right and left maxillary first molar teeth in an adult horse (14 years of age) in a retrospective study to describe clinical and CT findings for 74 horses and 7 ponies that underwent intraoral extraction of a cheek tooth with standing sedation and to assess potential associations between these features and procedure outcome (successful vs unsuccessful intraoral extraction). There is evidence of grade 3 and grade 2 infundibular caries in the rostral infundibula of the right and left maxillary first molar teeth, respectively (arrowheads), with moderate periapical sclerosis and widening of the periodontal spaces (arrows). The left rostral and caudal maxillary sinuses and ventral conchal sinus are filled with fluid, indicative of sinusitis (asterisks). L = Left. R = Right.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
Representative transverse CT image obtained at the level of the right and left maxillary first molar teeth in an adult horse (14 years of age) in a retrospective study to describe clinical and CT findings for 74 horses and 7 ponies that underwent intraoral extraction of a cheek tooth with standing sedation and to assess potential associations between these features and procedure outcome (successful vs unsuccessful intraoral extraction). There is evidence of grade 3 and grade 2 infundibular caries in the rostral infundibula of the right and left maxillary first molar teeth, respectively (arrowheads), with moderate periapical sclerosis and widening of the periodontal spaces (arrows). The left rostral and caudal maxillary sinuses and ventral conchal sinus are filled with fluid, indicative of sinusitis (asterisks). L = Left. R = Right.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
The presence of tooth root resorption (or blunting), fragmentation, or dental enlargement surrounding the apex of a root (typically described as a cementoma) was also recorded. A previously described grading system14 was used to grade infundibulae as normal (grade 0) or displaying hypoattenuation at the occlusal surface (grade 1); linear hypoattenuation along the majority of the infundibular length, representing lack of infundibular cementum (grade 2); linear hypoattenuation with a bulbous shape at the apical extent, representing lack of infundibular cementum with enamel defects (grade 3); or a fracture plane through the infundibulum (grade 4).
A dental fracture was defined as simple when the crown was fractured into 2 pieces and comminuted when > 2 pieces were identified. The fracture orientation was also assessed on CT, and the fractures were classified into 3 categories according to their configuration: sagittal, slab, and transverse. Buccal and lingual or palatal slab fractures were grouped together for analysis purposes. Excessive curvature or angulation of the roots was noted. Protuberant cheek tooth reserve crown morphology was recorded as present when the reserve crown had a dorsal plane cross-sectional area greater than, or an offset center relative to, the cross-sectional area of the clinical crown of the same tooth. If present, convergence of adjacent cheek teeth with the extracted tooth was noted when assessed in a sagittal or parasagittal plane on CT images. The maximum tooth length was measured on transverse reconstructed images from the central point of the occlusal surface to the central point between the root bases (Figure 2) The clinical crown length was measured on the palatal or lingual and buccal aspects of the tooth from the margin of the alveolar bone to the occlusal surface of the clinical crown, halfway between the mesial and distal surfaces. Thus, the clinical crown length determined by CT in the present study included a portion of the tooth located subgingivally that could not be measured on CT images. However, the authors believed that this CT measurement was practically relevant because purchase of the crown can be gained to the level of the alveolar bone margin following gingival elevation. The presence of sinusitis and the affected paranasal sinus compartments were noted; the predominant sinus affected was identified as the sinus with the largest volume of soft tissue (fluid)-attenuating material as determined by evaluation of the CT images.
Transverse CT image of the left maxillary first molar tooth of the horse in Figure 1 depicting measurements. Maximum tooth length was measured from the central point of the occlusal surface (white line) to the central point between the root bases (black line; B). Clinical crown length was measured on the palatal or lingual (black line; A) and buccal aspects (black line; C) of teeth from the margin of the alveolar bone to the occlusal surface of the clinical crown, halfway between the mesial and distal surfaces.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
Transverse CT image of the left maxillary first molar tooth of the horse in Figure 1 depicting measurements. Maximum tooth length was measured from the central point of the occlusal surface (white line) to the central point between the root bases (black line; B). Clinical crown length was measured on the palatal or lingual (black line; A) and buccal aspects (black line; C) of teeth from the margin of the alveolar bone to the occlusal surface of the clinical crown, halfway between the mesial and distal surfaces.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
Transverse CT image of the left maxillary first molar tooth of the horse in Figure 1 depicting measurements. Maximum tooth length was measured from the central point of the occlusal surface (white line) to the central point between the root bases (black line; B). Clinical crown length was measured on the palatal or lingual (black line; A) and buccal aspects (black line; C) of teeth from the margin of the alveolar bone to the occlusal surface of the clinical crown, halfway between the mesial and distal surfaces.
Citation: Journal of the American Veterinary Medical Association 255, 12; 10.2460/javma.255.12.1369
Intraoral extraction procedures
All intraoral extractions were performed by veterinary surgeons with board certification. A previously described technique3 was used, and the procedures were performed in standing, sedated horses and ponies unless conversion to a different approach was required. Maxillary and mandibular regional nerve blocks were used for all intraoral extractions, with approaches as previously reported in the literature.25 Angled dental picks, curettes, chisels, and gouges were used to complete intraoral extraction when required. Postextraction radiography or CT image acquisition were not routinely performed.
Statistical analysis
Clinical factors and CT findings assessed for association with outcome were selected on the basis of suspected causal relationships with the success of intraoral extraction. The variables included were demographic and clinical factors (age, breed, sex, and duration of clinical signs), general characteristics of teeth to be extracted (dimensions, type, and position), apical changes (sclerosis, root blunting, clubbing, or fragmentation), infundibular grades (for the maxillary cheek teeth), the presence and type of dental fractures, the integrity of the lamina dura, the appearance of periodontal space, presence of sinusitis, any protuberant crown morphology, convergence of the adjacent tooth crown, and abnormal root angulation. Descriptive data were generated for each variable. Comparisons were made between the teeth for which intraoral extraction was and was not successful. Each factor was first tested by univariable logistic regression by use of the deviance test and considered a candidate for multivariable analysis when the P value was ≤ 0.1. Statistical analysis was conducted with and without the individual animal as a random effect to account for multiple tooth extractions from the same equid. A forward selection procedure was applied for the model selection with the deviance test. The procedure was stopped once all covariates were contributing to the model with values of P ≤ 0.1. As a robustness check, the variables selected in the final model were tested to determine whether they could be deleted without significantly increasing the deviance value and to ensure that no term could be omitted. Finally, each factor excluded during initial model building was forced back into the final model. A coefficient change of > 30% was considered to be an indicator of significant confounding effect. All variables in the final logistic regression model were expressed with ORs (reflecting the odds of successful intraoral extraction) and 95% CIs, P values, and log likelihood P values (used to assess the model fit and contribution of each variable to the final selected model). For the final analyses, values of P ≤ 0.05 were considered significant. All data analyses were performed with standard statistical software.c
Results
Eighty-one equids that underwent intraoral extraction of 89 cheek teeth met the study inclusion criteria. The sample included 74 horses (30 warmblood type or warmblood crosses, 17 Thoroughbreds or Thoroughbred crosses, 12 cob type, and 15 horses of other mixed breeds or for which breed was unspecified) and 7 ponies (6 Welsh ponies or Welsh pony crosses and 1 Shetland pony [all grouped together as ponies for the purposes of statistical analysis]). No miniature breed horses were included. There were 46 males (57%; 45 geldings and 1 stallion) and 35 females (43%; all sexually intact). The median age at the time of the initial tooth extraction was 11 years (range, 2 to 24 years); for the 7 equids that underwent multiple extractions (5 that had 2 extractions each and 2 that had 3 extractions each), most procedures were done during the same visit and sedation episode. Four horses had 2 extractions and 1 had 3 extractions performed during a single visit. For the latter horse, the third cheek tooth that was removed was excluded from the study because it was adjacent to the site of one of the previously extracted teeth.
The main reason for initial evaluation at the referral hospital was unilateral nasal discharge in 36 of 81 (44%) patients, facial swelling of the mandibular or maxillary region in 16 (20%) patients, and nonspecific poor performance (including headshaking, problems taking the bit, riding, or abnormal head carriage) in 10 (12%) patients. Five patients (6%) were evaluated because of quidding or dysphagia, 2 (2%) had an external draining tract, and 1 (1%) had halitosis. For the remaining 11 (14%) patients, the CT examination and intraoral extraction were performed after a fractured tooth had been identified during oral examination by the referring veterinarian, and no other associated clinical signs had been noted. Of the 70 patients with detectable clinical signs, 53 (76%) had a known duration. Of these 53, 18 (34%) had a duration < 1 month, 24 (45%) had a duration of 1 to 6 months, and 11 (21%) had a duration > 6 months. Oral examination prior to CT was not routinely performed and was reported for 21 of 81 (26%) patients, including 14 with dental fracture. For these animals, all fractured teeth noted on CT evaluation had been identified during oral examination. The 2 patients that were admitted twice had recurrence of nasal discharge. The 89 teeth removed by intraoral extraction included 80 maxillary and 9 mandibular cheek teeth.
The targeted teeth were most commonly right or left maxillary (39/89 [44%]) or mandibular (2 [2%]) first molar teeth, followed by maxillary (11 [12%]) or mandibular (5 [6%]) third premolar, maxillary second molar (12 [13%]), and maxillary fourth premolar (9 [10%]) teeth. The remaining targeted teeth were right or left maxillary (4 [4%]) or mandibular (1 [1%]) second premolar and maxillary (4 [4%]) or mandibular (1 [1%]) third molar teeth. One (1%) was a maxillary supernumerary tooth. Three teeth were each adjacent to a tooth that had been previously removed and were included in the study because the time between extractions was > 3 months in all cases and CT was repeated before the second extraction. Sixty-two of 80 (78%) extracted maxillary teeth were associated with signs of ipsilateral sinusitis on CT, with the rostral maxillary sinus predominately involved (47/62 [76%]), followed by the ventral conchal sinus (9 [15%]), caudal maxillary sinus (5 [8%]), and dorsal conchal sinus (1 [2%]).
Intraoral extraction of 60 of 89 (67%) cheek teeth, including 56 of 80 (70%) maxillary and 4 of 9 mandibular cheek teeth, was deemed successful. Conversion to an alternative surgical extraction technique was elected for 20 maxillary and 5 mandibular cheek teeth. Small dental fragments were left in situ without any other surgical approach attempted after the intraoral extraction of 4 maxillary cheek teeth (causing these to be classified as unsuccessful extractions).
Seventeen maxillary cheek teeth required an MIB approach, and extraction was complete for 15 of these teeth; 1 of the remaining 2 was successfully repulsed, and small dental fragments were left in situ for 1. Repulsion was used without any other surgical approach for 3 maxillary cheek teeth and resulted in complete extraction for all 3. Of the 5 mandibular cheek teeth for which intraoral extraction was unsuccessful, MIB was attempted for 4 and resulted in complete extraction for 1; extraction was completed by repulsion of the remaining 3 mandibular cheek teeth. For 1 mandibular cheek tooth that had an unsuccessful intraoral extraction attempt, repulsion was used to complete the extraction without any other approach. All repulsions were performed under general anesthesia. Stabilization of fractured teeth with PMMA or partial coronectomy was not performed in any patients.
Logistic regression was performed with and without the horse as a random effect. Because only 7 of 81 horses and ponies had > 1 tooth extracted, logistic regression with the horse included as a random effect was not suitable, and the results obtained to define the final model were not reliable. As a result, the final model was chosen without including the horse as a random effect (ie, multiple extractions were treated as independent).
The presence, type, and orientation of a dental fracture, clubbing of the apex of the distal root, clubbing of the apex of the palatal root (for maxillary teeth), and the functional sinus compartment affected by sinusitis were each associated with lower odds of successful intraoral extraction on univariable analysis (Table 1). The number of clubbed roots on the targeted tooth (mean ± SD, 1.0 ± 1.0 for teeth that had unsuccessful intraoral extraction [n = 29] vs 0.5 ± 0.7 for teeth that had successful intraoral extraction [60]) was also associated with this outcome on univariable analysis (OR, 0.52 [95% CI, 0.31 to 0.87]; P = 0.014; log likelihood, P = 0.011).
Results of univariable analysis for factors potentially associated with successful (vs unsuccessful) intraoral extraction of 89 cheek teeth in 8I horses and ponies in a retrospective study.
| Variable | No. of teeth | Category | Unsuccessful intraoral extraction (No. [%]) | Successful intraoral extraction (No. [%]) | OR (95% CI) | P value | Log likelihood P value |
|---|---|---|---|---|---|---|---|
| Fracture | 89 | Absent | 12 (21) | 46 (79) | Referent | — | 0.001 |
| Present | 17 (55) | 14 (45) | 0.21 (0.08–0.56) | 0.002 | |||
| Fracture type | 89 | No fracture | 12 (21) | 46 (79) | Referent | — | < 0.001 |
| Simple | 12 (46) | 14 (54) | 0.30 (0.11–0.83) | 0.02 | |||
| Fracture orientation | 89 | Absent | 12 (21) | 46 (79) | Referent | — | 0.005 |
| Midline sagittal | 5 (56) | 4 (44) | 0.21 (0.05–0.90) | 0.035 | |||
| Buccal or palatal slab | 12 (55) | 10 (45) | 0.22 (0.08–0.62) | 0.004 | |||
| Clubbing of palatal root* | 80 | Absent | 16 (24) | 50 (76) | Referent | — | 0.019 |
| Present | 8 (57) | 6 (43) | 0.24 (0.07–0.89) | 0.02 | |||
| Clubbing of distal root | 89 | Absent | 19 (26) | 53 (74) | Referent | — | 0.013 |
| Present | 10 (59) | 7 (41) | 0.25 (0.08–0.75) | 0.014 | |||
| Functional sinus | 80 | None | 3 (17) | 15 (83) | Referent | — | 0.028 |
| compartment involved* | Rostral | 12 (55) | 10 (45) | 0.17 (0.04–0.74) | 0.011 | ||
| Rostral and caudal | 9 (23) | 30 (77) | 0.67 (0.16–2.83) | 0.58 |
Only results for variables significantly associated with the outcome on univariable analysis are shown. The ORs represent the odds of successful intraoral extraction for a tooth in a given category, relative to the referent category. Fracture type included 5 teeth with comminuted fractures for which intraoral extraction was unsuccessful, and functional sinus compartment involvement included 1 tooth with caudal sinus compartment involvement for which intraoral extraction was successful; these data were excluded from the analysis because of small numbers.
Applicable to maxillary teeth only.
— = Not applicable.
Age, breed, sex, main reason for evaluation, and duration of clinical signs were not significantly associated with this outcome on univariable screening. Among the CT features, the variables excluded after the univariable analysis include the tooth position on the arcades, the type of tooth (maxillary or mandibular), the tooth dimensions (length of buccal or palatal margins of the clinical crown as well as the maximum tooth length), the presence of gas in the pulp cavity, periodontal space widening, abnormal or absent lamina dura, periapical sclerosis, and convergence of adjacent clinical crowns.
Infundibular grades, the presence of protuberant cheek tooth reserve crown morphology, convergence of adjacent cheek teeth with the extracted tooth, blunting of the palatal root, fragmentation of the caudal root, and the duration of clinical signs were identified statistically as potential confounders. Given the limited observations and insufficient repetitiveness, 3 of these variables (the infundibular grades, fragmentation of the caudal roots, and presence of protuberant cheek tooth reserve crown morphology) had large CIs, and most of the factors had a high SE. As a result, selection of this model was not considered reliable to draw further statistical conclusions, and exclusion of the confounding variables was elected in the final multivariable model. On multivariable analysis, presence of a simple fracture was the only factor associated with the outcome of interest; the odds of successful intraoral extraction were significantly (P = 0.02) lower for teeth that had this finding, compared with those for teeth that had no fracture (OR, 0.30; 95% CI, 0.13 to 0.70). Of the 26 cheek teeth with simple fractures, 14 (54%) had successful intraoral extraction. Cheek teeth with a simple slab fracture predominated (20/26 [77%]), followed by simple sagittal fracture (6/26 [23%]); intraoral extraction was successful for 10 of 20 and 4 of 6 teeth with simple slab and sagittal fractures, respectively. No teeth in the study had transverse fractures, and the number of simple sagittal fractures was insufficient for detailed comparison between the 2 fracture configurations. The fracture planes penetrated the pulp in all fractured teeth identified in the study, but the presence of gas in the pulp cavity was not retained during logistic regression (excluded after univariable screening). Palatal root clubbing, although included in the final multivariable model, was not significantly (P = 0.093) associated with the outcome of intraoral extraction in this analysis.
Discussion
Periapical tooth root infection can manifest in several different ways depending on the severity and chronicity. Detection of intrapulpar gas by CT in equine patients with tooth root infection has been described in 14 of 28 (50%) to 16 of 18 cases in other studies.14,15 Other important CT indicators of dental disease include alveolar bone sclerosis, root clubbing, blunting and fragmentation, periapical gas, widening of the periodontal space, and mucosal thickening of the overlying paranasal sinuses.14,15,18,21
The success rate of intraoral extraction in the sample of the present study (60/89 [67%]) was lower than previously reported.3,4,6,9 It is worth mentioning that intraoral extraction was attempted first for all patients in this study and was considered unsuccessful even if a very small portion of a root remained in situ; this approach and the stringent cutoff used may have contributed to the apparently lower success rate in our study. In 5 of 9 mandibular cheek tooth extractions, intraoral extraction was unsuccessful, compared with 24 of 80 (30%) maxillary cheek tooth extractions. Our experience suggests that mandibular exodontia is more challenging. The cheek teeth are not as large in the mandible as they are in the maxilla, and the jaw is also narrower, making the placement of instruments and the mechanics of extraction more difficult. Interestingly, tooth position on the mandibular arcade was not associated with lower odds of successful intraoral extraction in this study. Intraoral extraction is often considered more demanding in young equids in which the reserve crown is markedly elongated and has strong periodontal attachments.5,18 The horses and ponies of the present study were predominantly middle-aged, and no association between patient age or maximum tooth length and the success of intraoral extraction was identified.
Breed and sex were excluded during the statistical model building process. No horses from miniature breeds were included in the present study, and some potentially breed-related factors such as oral cavity size or the degree of maximal mouth opening were not investigated for associations with outcome. Patient temperament was also not evaluated; given that all intraoral extractions in the study were performed in standing sedated horses and ponies, it is possible that extraction under general anesthesia would have facilitated intraoral extraction in some less tolerant patients. Evidence supporting that distally located teeth may be more challenging to remove because of a lack of space1 was not found in our study.
Results of previous studies4,8,10 indicate that intraoral extraction of cheek teeth with preexisting fractures can be successful without employing more invasive approaches. However, extraction of such teeth is usually more challenging.4,7 The results of the final logistic regression model in the present study supported these observations, as cheek teeth with a simple fracture had significantly lower odds of successful intraoral extraction than nonfractured teeth. Most of the horses and ponies included in the study were initially referred to our facility for CT evaluation of the head, and oral examinations were not routinely performed prior to imaging. Common dental fractures are usually straightforward to identify on oral examination without the need for CT, and this was true for the 14 patients of the present study with fractures that underwent oral examination, although CT allows for assessment of the apices and other associated structures. Tooth position, oral access, crown size, and fracture configuration may contribute to marked differences in the inherent difficulty of fractured cheek tooth removal. Whereas Ramzan et al8 reported successful intraoral extraction of 27 of 31 fractured cheek teeth under endoscopic guidance, midline sagittal fractures of maxillary teeth (13/31 [42%]) were the most common type, followed by slab fractures (10 [32%]), and transverse or comminuted fractures (5 [16%]). In the present study, intraoral extraction was successful for 14 of 31 fractured cheek teeth, and simple slab fractures (n = 20 [65%]) were the most frequent, followed by simple sagittal fractures (6 [19%]) and comminuted fractures (5 [16%]). Our results supported a previous observation that slab fractures are more common than sagittal fractures.18 When looking at the characteristics of the simple fractures, it might be considered that teeth with slab fractures may be more challenging to extract than those with sagittal fractures (with successful intraoral extraction for 10 of 20 and 4 of 6, respectively, in the present study), notably because slab fractures reduce the area of contact for molar spreaders and forceps on the clinical crown.1 This was not supported by the statistical analyses in our study, perhaps because of the small sample size. To the authors' knowledge, there is no current evidence that reconstruction of the clinical crown by use of PMMA in equids with a slab fracture improves the success of intraoral extraction. For sagittal fractures, the displacement and depth of the fracture plane were not measured in our study, and it was unknown whether stabilizing the fragments with PMMA would have improved the outcome. The 5 cheek teeth with comminuted dental fractures included in the present study were clinically observed to be the most challenging to extract orally, but the number of cases was too small to determine whether there was any significant association between this fracture configuration and the success of intraoral extraction.
All fractured teeth in the study reported here had evident pulp exposure. However, this feature alone is not definite evidence of pulp necrosis and active periapical infection, a phenomenon likely to be explained by sealing of affected pulp horns with tertiary dentin.26–28 Even if a bridge of tertiary dentin is present apically, exposure of more brittle material, such as the dentin present around the pulp, is likely to compromise clinical crown integrity when the tooth is gripped with molar forceps. However, our multivariable analysis revealed no significant association between the presence of any fracture (and thus pulp exposure) or the presence of gas in the pulp cavity and successful versus unsuccessful intraoral extraction.
Widening of the periodontal space, indicating disruption of the periodontal ligament, may increase the likelihood of a successful intraoral extraction12,15,16,21; however, this effect was not identified in the present study. The normal lamina dura is a very thin bone structure in close proximity to the tooth roots and the unerupted portion of the tooth. As a solitary feature, a nondetectable lamina dura should be interpreted cautiously (as different imaging systems may have various spatial and contrast resolutions) and may be considered normal in some cases.14 An abnormal or nonvisible lamina dura on CT is unlikely to be associated with intraoral extraction success, and no association between the presence or absence of lamina dura and the success of the intraoral extraction was evident in the study reported here. We considered it likely that identification of root clubbing could be a pertinent CT finding, as it could predispose the roots to fracture in the alveolar socket during extraction and thus being left in situ,4,5 particularly in older equids. However, clubbing of the palatal and distal roots and the number of clubbed roots, although associated with lower odds of successful intraoral extraction in the univariable analysis, were not significantly associated with this outcome in the multivariable analysis.
Gas within the infundibulum, hypoattenuation of the cementum, and destruction of the enamel are the most frequent CT findings associated with infundibular caries,14 and the mesial infundibulum is most frequently affected.18 A sagittal fracture plane commonly encompasses mesial or distal infundibula (or both), suggesting that severe infundibular caries could have a role in unsuccessful intraoral extraction by predisposing the crown to disintegrate.4,11,16,17 Unfortunately in the present study, the infundibular grades were excluded from the analysis because of confounding and could not support this association.
Cheek teeth with small or partially absent clinical crowns offer insufficient surfaces for contact with molar forceps, making manipulation and intraoral extraction more difficult or potentially impossible.1 Statistical analysis did not identify an association between the length of the buccal or palatal margins of the clinical crown and successful versus unsuccessful intraoral extraction in the present study. There could be a correlation between the volume of missing clinical crown and success of intraoral extraction; however, we did not assess the percentage of clinical crown loss at the occlusal surface, and it may be beneficial to assess this feature in future studies. Protuberant cheek tooth reserve crown morphology and convergence of adjacent clinical crowns were previously identified as potential risk factors for failure of the minimally invasive transbuccal screw extraction technique.5 In the study reported here, each of these factors had limited observations. The convergence of the adjacent clinical crowns was not retained after univariable screening, and the protuberant cheek tooth reserve crown morphology factor was excluded from the statistical analysis because of confounding.
Despite including 89 extracted cheek teeth, the authors acknowledge that identification and analysis of factors associated with the outcome of intraoral extraction in equids in the study reported here could have been limited by the lack of sufficient data for some variables (eg, protuberant cheek tooth reserve crown morphology or convergence of adjacent clinical crowns). This lack of observations also resulted in the selection of a final multivariable model that excluded potentially confounding factors (with large CIs and high SEs). A decision was made to select a multivariable model without controlling for the individual horse as a random effect and to consider each of the teeth independently. Only 7 of 81 horses had > 1 tooth extracted, and the results were not reliable when the analysis was performed with the individual as a random effect. Statistically, it was possible that the removal of the horses with multiple extractions would have influenced the final results. However, from a clinical standpoint in these cases, the authors were of the opinion that the biological effect of the individual animal was minimal, especially with the exclusion criteria used for multiple tooth extraction. With a larger group of horses and ponies, additional features that are significantly associated with the outcome of this procedure may be identified, and this could form the basis for additional, possibly multi-institutional, investigations.
Results obtained in the present study included clinically useful data that can be used by practitioners to refine cheek tooth extraction plans for horses and ponies. These findings can be used to inform owners regarding the risks of managing cheek teeth, particularly those with simple fractures, by routine intraoral extraction and the potential requirement for extraoral surgical approaches.
Acknowledgments
The authors did not receive any third-party funding or support in connection with this study or the writing of the manuscript and declare that there were no conflicts of interest.
The authors thank Chen Xianyan, University of Georgia, for statistical analysis.
ABBREVIATIONS
| CI | Confidence interval |
| MIB | Minimally invasive buccotomy |
| PMMA | Polymethyl methacrylate |
Footnotes
GE Lightspeed Pro 16, GE Medical Systems, Buckinghamshire, England.
OsiriX 64-bit, version 6.0.2, Pixmeo Sàrl, Bernex, Switzerland.
R, version 3.2.0, R Foundation for Statistical Computing, Vienna, Austria. Available at: www.R-project.org/. Accessed Aug 17, 2018.
References
1. Lowder MQ. How to perform oral extraction of equine cheek teeth, in Proceedings. 44th Annu Meet Am Assoc Equine Pract 1999;131–136.
2. Dixon PM, Tremaine WH, Pickles K, et al. Equine dental disease part 4: a long-term study of 400 cases: apical infections of cheek teeth. Equine Vet J 2000;32:182–194.
3. Tremaine WH. Oral extraction of equine cheek teeth. Equine Vet Educ 2004;16:151–158.
4. Dixon PM, Dacre I, Dacre K, et al. Standing oral extraction of cheek teeth in 100 horses (1998–2003). Equine Vet J 2005;37:105–112.
5. Langeneckert F, Witte T, Schellenberger F, et al. Cheek tooth extraction via a minimally invasive transbuccal approach and intradental screw placement in 54 equids. Vet Surg 2015;44:1012–1020.
6. Prichard MA, Hackett RP, Erb HN. Long-term outcome of tooth repulsion in horses. A retrospective study of 61 cases. Vet Surg 1992;21:145–149.
7. O'Neill HD, Boussauw B, Bladon BM, et al. Extraction of cheek teeth using a lateral buccotomy approach in 114 horses (1999–2009). Equine Vet J 2011;43:348–353.
8. Ramzan PH, Dallas RS, Palmer L. Extraction of fractured cheek teeth under oral endoscopic guidance in standing horses. Vet Surg 2011;40:586–589.
9. Rice MK, Henry TJ. Standing intraoral extractions of cheek teeth aided by partial crown removal in 165 horses (2010–2016). Equine Vet J 2018;50:48–53.
10. Pearce DJ, Brown JA. Extraction of 22 equine cheek teeth with displaced sagittal fractures using polymethylmethacrylate stabilisation (2011–2016). Equine Vet Educ 2017;31:421–426.
11. Townsend NB, Hawkes CS, Rex R, et al. Investigation of the sensitivity and specificity of radiological signs for diagnosis of periapical infection of equine cheek teeth. Equine Vet J 2011;43:170–178.
12. Weller R, Livesey L, Maierl J, et al. Comparison of radiography and scintigraphy in the diagnosis of dental disorders in the horse. Equine Vet J 2001;33:49–58.
13. Gibbs C, Lane JG. Radiographic examination of the facial, nasal and paranasal sinus regions of the horse. II. Radiological findings. Equine Vet J 1987;19:474–482.
14. Bühler M, Fürst A, Lewis FI, et al. Computed tomographic features of apical infection of equine maxillary cheek teeth: a retrospective study of 49 horses. Equine Vet J 2014;46:468–473.
15. Henninger W, Frame EM, Willmann M, et al. CT features of alveolitis and sinusitis in horses. Vet Radiol Ultrasound 2003;44:269–276.
16. Veraa S, Voorhout G, Klein WR. Computed tomography of the upper cheek teeth in horses with infundibular changes and apical infection. Equine Vet J 2009;41:872–876.
17. Selberg K, Easley JT. Advanced imaging in equine dental disease. Vet Clin North Am Equine Pract 2013;29:397–409.
18. Barakzai SZ, Barnett TP. Computed tomography and scintigraphy for evaluation of dental disease in the horse. Equine Vet Educ 2015;27:323–331.
19. Windley Z, Weller R, Tremaine WH, et al. Two- and three-dimensional computed tomographic anatomy of the enamel, infundibulae and pulp of 126 equine cheek teeth. Part 1: findings in teeth without macroscopic occlusal or computed tomographic lesions. Equine Vet J 2009;41:433–440.
20. Windley Z, Weller R, Tremaine WH, et al. Two- and three-dimensional computed tomographic anatomy of the enamel, infundibulae and pulp of 126 equine cheek teeth. Part 2: findings in teeth with macroscopic occlusal or computed tomographic lesions. Equine Vet J 2009;41:441–447.
21. Liuti T, Smith S, Dixon PM. Radiographic, computed tomographic, gross pathological and histological findings with suspected apical infection in 32 equine maxillary cheek teeth (2012–2015). Equine Vet J 2018;50:41–47.
22. Liuti T, Reardon R, Dixon PM. Computed tomographic assessment of equine maxillary cheek teeth anatomical relationships, and paranasal sinus volumes. Vet Rec 2017;181:452.
23. Manso-Díaz G, García-López JM, Maranda L, et al. The role of head computed tomography in equine practice. Equine Vet Educ 2015;27:136–145.
24. Dakin SG, Lam R, Rees E, et al. Technical set-up and radiation exposure for standing computed tomography of the equine head. Equine Vet Educ 2014;26:208–215.
25. Tremaine WH. Local analgesic techniques for the equine head. Equine Vet Educ 2007;19:495–503.
26. Dacre I, Kempsot S, Dixon PM. Equine idiopathic cheek teeth fractures. Part 1: pathological studies on 35 fractured cheek teeth. Equine Vet J 2007;39:310–318.
27. Taylor L, Dixon PM. Equine idiopathic cheek teeth fractures: part 2: a practice-based survey of 147 affected horses in Britain and Ireland. Equine Vet J 2007;39:322–326.
28. Dixon PM, Barakzai SZ, Collins NM, et al. Equine idiopathic cheek teeth fractures: part 3: a hospital-based survey of 68 referred horses (1999–2005). Equine Vet J 2007;39:327–332.
