The present study was approved by the Institutional Review Board at Seoul National University Dental Hospital and carried out according to institutional guidelines. We identified a total of 334 patients diagnosed histologically with keratocystic odontogenic tumor from 2004 to 2010 at Seoul National University Dental Hospital. Among them, the patients that had undergone surgical treatment for KCOT were initially eligible for this study. Inclusion criteria were patients with a definite histopathologic diagnosis of KCOT who had a CT scan (Somatom Sensation 10, Siemens, Erlangen, Germany) taken during the preoperative analysis. Exclusion criteria were 1) patients treated elsewhere with no surgical treatment provided at SNUDH and 2) patients undergoing surgery due to recurrence.

Based on these criteria, in the end, 180 patients were included in the present study. Using their dental records, the available information was reviewed, which was gender, age at diagnosis, clinical manifestations, site of involvement, radiographic measurement, and treatment modality. KCOTs were diagnosed in one of two different ways. First, the lesion was identified incidentally by routine examination or radiograph. Second, it was clinically identified with symptoms such as pain, swelling or drainage.

The location of the KCOTs in the maxilla or mandible was noted, and they were further subdivided into the following groups: 1) maxillary incisor and canine, 2) maxillary premolar, 3) maxillary molar and tuberosity, 4) maxillary sinus, 5) mandibular incisor and canine, 6) mandibular premolar, 7) mandibular molar, 8) mandibular angle and ramus, 9) mandibular coronoid process, and 10) mandibular condyle.

The treatment provided for each KCOT was also recorded as determined by a review of the surgical report. All of the KCOTs described in this report were treated by enucleation with peripheral ostectomy. Any adjunctive method such as marsupialization, bone graft, or en bloc resection was noted with each case.

The radiographs were saved as digital imaging and communications in medicine (DICOM) files. All of the radiographs were evaluated by one radiologist.

The size of each KCOT was measured on the CT slices using the axial plane in millimeters. The measurements were made by using a picture archiving and communication system (PACS) viewer (PiView STAR, Infinitt, Seoul, Korea). On the axial view, the longest mesio-distal and bucco-lingual dimensions were obtained. For the mandibular lesions, the mesio-distal dimension was measured parallel to the long axis of the mandibular body (Fig. 1A). The bucco-lingual dimension was measured perpendicular to the long axis of the mandibular body (Fig. 1B). On the maxillary axial view, the mesio-distal dimension was measured parallel to the coronal plane of the maxilla (Fig. 2A), and the bucco-lingual dimension was measured parallel to the sagittal plane of the maxilla (Fig. 2B). If the lesion involved both the tooth bearing area and non-tooth bearing area, the size of the tooth bearing area was measured separately. The tooth bearing area was defined as the area mesial to the distal surface of the secondary molar and superior to the apex of the roots in the maxilla. If the lesion fenestrated the buccal or lingual bone wall, it was recorded as having cortical discontinuity.

The number of teeth removed during surgery and teeth treated with a root canal associated with the lesion were also counted. In cases where tooth removal was conducted during surgery, it was assumed that rehabilitation by means of dental implants was needed. Dental expenses were estimated as a sum total of all the medical costs of adjunctive treatment, root canal treatment, and the prosthetic rehabilitation of the missing tooth. The values were adjusted by the SNUDH price index. In order to perform statistical analysis, estimated dental expenses were then classified into 4 groups according to the values. The dental expenses increased from group1 to group4.

All of the measurements were evaluated for the maxillary and mandibular lesions separately. In cases of multiple lesions occurring independently in a separate jawbone, the lesion largest in size was subject to measurement.

All of the statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL, USA) for Windows version 16.0. A paired t-test and Pearson's χ² were used to assess the differences in patients with or without preoperative symptoms and cortical discontinuity in the study group. Analysis of variance (ANOVA) and Tukey's post hoc test were used for assessing the KCOT size by comparing several variables such as dental expenses, number of root canal treatments, and number of teeth removed. A 95% significance level was used for all of the statistical tests (P<0.05). Due to the homogeneity in the data within the study groups and a non-normal distribution of the data, all variables were transformed in order to reduce the variance and stabilize normality.

Our patient group was distributed in a wide age range from 9 to 80 years of age (mean, 36.8 years). There were 93 males (51.6%) and 87 females (48.3%) (ratio 1.1 : 1). The 117 patients (65%) presented with symptoms such as swelling, pain, drainage, and discomfort, whereas the remaining 63 (35%) were diagnosed incidentally at the time of routine examination (Table 1).

Cortical discontinuity was found in 119 of the 198 lesions (60.1%). Fifty-five lesions (27.7%) were maxillary and 143 were mandibular (ratio 1 : 2.6). The most frequently involved regions were the mandibular molar, angle, and ramus (57.2%). The maxillary molar and tuberosity region was the next most common location (11.0%). The majority (88.1%) of the KCOT's were located posterior to the canines (Table 2).

Radiographically, the size of the mandibular lesions ranged from 7.4 to 75.2mm(mean size of 29.9 mm) mesiodistally and from 4.9 to 27.5 mm (mean size of 14.0 mm) bucco-lingually. In the maxilla, the mean size of the lesions was 23.1 mm, with the smallest being 5.7 and the largest 44.1 mm mesio-distally, and the mean size of the lesion was 20.1 mm, with the smallest being 4.1 mm and the largest 49.2 mm. Most lesions measured over 10 mm (Figs. 3, 4, 5, 6). No statistically significant correlation with the KCOT size was found for age or for gender. Marsupialization was performed to treat 10% (18/180) of the cases. Bone graft was performed to treat 25% (45/180). Resection was performed to treat 1.1% (2/180) (Table 3).

All of the results are summarized in Tables 4, 5, 6. In the mandible, there were no statistically significant differences in bucco-lingual dimensions (P=0.218), number of teeth removed (P=0.335), number of root canal treatments (P=0.288), or dental expenses (P=0.239) between symptomatic patients and non-symptomatic patients. However, there were significantly smaller mesio-distal dimensions in non-symptomatic patients (P=0.016). In the maxilla, no statistically significant correlation was found between the initial symptoms and mesio-distal dimensions (P=0.182), bucco-lingual dimensions (P=0.689), number of teeth removed, number of root canal treatments, or dental expenses.

A statistically significant correlation was found for the presenting cortical discontinuity and the mesio-distal and bucco-lingual dimensions of the KCOTs (P<0.05, Table 6) in the mandible. In the maxilla, however, no statistically significant correlation was found between the cortical discontinuity and the mesio-distal (P=0.384) and bucco-lingual (P=0.682) dimensions of the KCOTs.

An ANOVA test for the dental expense data indicated that larger mandibular KCOT size was correlated with greater dental expenses (Table 7). Another ANOVA test indicated that the bucco-lingual dimension in tooth bearing area but not the mesio-distal dimension in the tooth bearing area increased with an increasing number of teeth removed (Table 8). In the maxillary KCOTs in the present study, there were no correlations between the size effects and the variables of interest. However, a statistically significant correlation was found between the bucco-lingual dimensions and dental expenses.

Finally, an ANOVA test indicated that the number of root canal treatments increased with the mesio-distal dimensions but did not differ by the bucco-lingual dimensions. In the maxilla, the number of root canal treatments increased with the mesio-distal and bucco-lingual dimensions of KCOTs (Table 9).