Dental implants are gaining popularity in the Middle East. According to recent surveys in the Middle East, 76% to 80% of respondents with tooth loss were willing to undergo implant therapy ¹ . Newer surgical techniques have allowed placement of dental implants in clinical conditions which were previously contraindicated ² . The lateral window technique in sinus lift surgery is sometimes associated with risk of hemorrhage as a result of inadvertent laceration of the lateral intraosseous blood vessels located in the lateral wall of the maxillary antrum^3,4. Due to their intraosseous location in the lateral antral wall, these vessels have an estimated 20% chance of sustaining injury during direct sinus lift procedures ^5,6 . Recently published studies have revealed variations in occurrence and location of the lateral antral intraosseous canals (LAIC) ^4,7-11. Currently, no studies regarding the incidence and location of the lateral antral intraosseous blood vessels in the Middle Eastern population have been published. We conducted a study to evaluate the occurrence and location of lateral antral intraosseous vascular canals in a United Arab Emirates (UAE) population.

The study was conducted between January to August 2020. Cone-beam computed tomography (CBCT) images from 200 patients (400 maxillary sinuses) at Thumbay Dental Hospital, Gulf Medical University, Ajman, UAE, were evaluated in this retrospective study. Prior to commencement of the study, ethical clearance was obtained from the Institutional Review Board of Gulf Medical University (No. INT/COD/FR/006-2020), and the written informed consent was obtained from all patients. The CBCT examinations of the sino-nasal variants were performed using a ProMax 3D Mid (Planmeca, Helsinki, Finland), operated at 90 kVp and 10 mA with a 9×16 cm field of view. Assessment of CBCT scans was directly performed on a 1,920×1,080 pixel and 23-inch DELL monitor (DELL, Round Rock, TX, USA). Voxel edge length was 0.2 mm, and slice thickness was 1 mm. The CBCT scans of patients with a history of trauma or surgery involving the maxillofacial region or those with clinical or radiographic indication of developmental anomalies affecting the maxillofacial region were excluded from the study. The CBCT scans of subjects with first premolars located anterior to the maxillary antrum were excluded from the study, following research by Mung San Aung et al. ⁴ . The CBCT scans of the subjects were evaluated by an oral radiologist and an oral surgeon with 10 years of clinical experience. In cases of inter-observer disagreement, a 3rd examiner (an oral surgeon with 10 years of clinical experience) was consulted for the final decision. To determine intra-observer variance, 50 CBCT scans (100 maxillary antra) were examined twice with one-month by the same examiners (r=0.92).

The criteria used for identification and dimensional measurement of LAIC were from the study conducted by Mardinger et al. ¹² . According to these criteria, four reference CBCT landmarks were considered for locating the position of the LAIC.(Fig. 1) The reference points were the centres of the crowns of the first premolar, second premolar, first molar, and second molar mesiodistally and bucco-lingually at the level of the alveolar crest in the axial view, with an equivalent distance from the adjacent tooth of the dentate side ¹³ . The diameters of the LAIC were divided into four categories based on the criteria employed by Mardinger et al. ¹² in their computed tomography (CT)-based study. To locate the reference points in the sub-sinusal edentulous maxilla (Fig. 2) and determine the horizontal dimensions of LAIC, the technique devised by Lee et al. ¹³ was applied. The data were analysed using IBM SPSS Statistics (ver. 26; IBM, Armonk, NY, USA). A P-value <0.05 was considered statistically significant.

The prevalence of LAIC was 62.3% (249 maxillary antra) among the study population (200 subjects, 400 maxillary antra). Among the 200 CBCT scans, 137 scans (274 maxillary antra) were of male subjects, and 63 scans (126 maxillary antra) were of female subjects. The prevalence of LAIC in the maxillary antra of male subjects was 63.1% (173 maxillary sinuses). The prevalence of LAIC in the maxillary antra of female subjects was 60.3% (76 maxillary antra). There was no statistically significant difference in prevalence of LAIC between male and female subjects (P=0.542). The mean distance between the LAIC and the alveolar crest in the posterior maxillary teeth is stated in Table 1. The mean distance between the most inferior point of the alveolar bone and the inferior border of the LAIC in the posterior maxillary region was 19.83±3.12 mm.

There was a significant difference between maxillary molar and premolar regions in mean distance from the most inferior point of the alveolar bone and the inferior border of the LAIC.(Table 2)

Fifty-three subjects had an edentulous area in the maxillary or mandibular region. Among them, 32 subjects had a sub-sinusal edentulous area (edentulous area extending anteriorly from the mesial aspect of the maxillary first premolar posteriorly to the distal aspect of the maxillary second molar). Among these 32 subjects, 12 had bilateral sub-sinusal edentulous areas and 20 had a unilateral edentulous area. There was no statistically significant difference in mean distance from the most inferior point of the alveolar bone and the inferior border of the LAIC between dentulous and edentulous areas (P=0.1).(Table 3)

The LAICs were categorised according to medio-lateral location within the antral wall based on the criteria by Lee et al. ¹³ .(Fig. 3) In only 3 cases, the canal bulged towards the outer side of the wall (G1-extrasinusal).(Fig. 4. A) Canals were embedded within the wall (G2-intraosseous) (Fig. 4. B) in 70 sinus walls, whereas canals bulged towards the inner side of the wall in 176 sinus walls (G3-intrasinusal).(Fig. 4. C) The diameter of the antral canal was classified based on the criteria devised by Mardinger et al. ¹² . The diameter was less than 1 mm in the majority of cases (n=113). Eighty-three LAICs had a diameter of 1-2 mm, and 53 LAICS had a diameter of 2-3 mm.(Fig. 5)

In the present study, LAIC was detected in 62.3% of 200 CBCT scans. A similar percentage (67.1%) was reported by Rahpeyma et al. ⁷ , although their study only evaluated 35 CBCT scans. In a Thai study of 184 CBCT scans, the LAIC was detected in 60.3% of the samples ⁴ . Higher percentage was reported in recently conducted studies in a Chinese population ¹⁴ (87.6% of 484 CBCT scans) and a Mexican population ¹⁵ (90% of 1,005 CBCT scans). One of the earliest reported studies ¹² in 2007 demonstrated lateral intraosseous antral canals in 55% of 208 scans. However, the study was conducted using CT. Another CT based study conducted in 2016 reported presence of intraosseus antral canal in 32% of the 284 scans that were examined ¹³ . Generally, CBCT-based studies show a higher percentage and better detection of LAIC than CT-based studies, perhaps because CBCT tends to display a superior spatial resolution compared to fan beam computed tomography ¹⁶ . In addition, some researchers have suggested that the higher spatial resolution observed in CBCT is due to the smaller voxel size of the scan ¹⁷ .

On the other hand, the differences in percentages of LAIC could be attributed to racial variations, since these studies were carried out in different parts of the world.

The present study demonstrated no significant difference in prevalence of LAIC between male and female subjects. Similar outcomes were reported by Lee et al. ¹³ and Rahpeyma et al. ⁷ . In one recent CBCT-based study, the occurrence of lateral antral canal walls was higher in males compared to females ⁶ . The researchers stated that the probability of detection of antral canals increases significantly as thickness of the antral wall increases, which is the probable reason for the increased radiographic detection of antral canals in male subjects in their study. When the location of the canal in the mediolateral aspect of the antral wall was evaluated, we found that 78% of the canals had an intrasinusal position. Similar radiographic features were reported by Lee et al. ¹³ . Extrasinusal is the least common location of the canal within the antral wall. However, one Korean study ¹⁸ conducted on cadavers reported that the extrasinusal variety was most common.

When the diameter of LAIC was evaluated, most were less than 1 mm. The diameter of the canals plays a very important role in the radiographic detection of antral canals. LAICs with a diameter less than 0.5 mm are difficult to detect on routine CBCT scans ¹⁹ . Furthermore, the diameter of the LAIC has clinical significance in surgical interventions in the posterior maxillary sinus antrum area as its alerts the surgeon to possible risk of hemorrhage during surgery ²⁰ . The surgeon can adjust the location of the osteotomy window according to pre-surgical CBCT evaluation to avoid complications ⁹ . Despite a detailed radiographic examination of the surgical site, however, accidental exposure of the antral vessels, which might lead to bleeding, can occur ^20,21 . This can be attributed to the heterogeneous pathways of the blood vessels, which can be partially intraosseous or completely extraosseous ²² . Researchers have advised the use of electrocautery or bone wax for management of bleeding subsequent to accidental rupture of these blood vessels ²² . Future CBCT-based research should focus on not only determining the point position of the LAIC, but also on tracing its pathways.

This study established the average measurement for location of the LAIC with reference to maxillary alveolar bone in the UAE, which has not been previously reported. This data will aid the surgeon in avoiding the LAIC during sinus lift procedures and reduce the risk of hemorrhage.