A Study of the Septa in the Maxillary Sinuses and the Subantral Alveolar Processes in 30 Patients

The septa of the maxillary sinuses in humans are thin walls of bone inside the sinus. They vary in number, thickness, and length, and they play a role in determining the placement of implants in the posterior maxillar.1,^–5 The incidence of these septa in computerized tomography (CT) was found to occur in 13% to 36% of the sinuses in a group of 30 patients.1^,3^,5,^–10 When studying the incidence of the septa in panoramic radiography, the septa of this same group could not be observed.5^,7 Confirmation of these findings was established during the surgery.

In studies by Bolin and Eliasson,11 the height of the alveolar process was compared using panoramic radiography and CT scans; they concluded that the posterior sectors of the maxilla demonstrated the least amount of distortion.

The objective of this study was to record the incidence of sinus septa and the length of the subantral alveolar processes and to compare the data obtained from the panoramic radiography and maxillary CT scans.

The study included 30 patients, 18 men and 12 women, ranging in age from 37 to 82 years. All patients required treatment with dental implants. The maxillary sinuses observed in panoramic radiographs and CT scans were classified as group 1 (totally edentulous) and group 2 (partially edentulous). The panoramic radiographs were carried out using the Panelipse II (General Electric Medical System, Milwaukee, Wis) and the CT scans with the XRI, which produced helicoidally 4 cuts/segment (General Electric Medical System) using the Dentascan program (General Electric Medical System).

Two common reference points were determined in the panoramic radiographs and the CT scans: (A) the vertex created by the pyriform apertures and the frontal maxillary apophysis on the right sides, and (B) the same landmarks on the left sides. A transverse straight line was traced joining the 2 anterior points until the posterior wall of the maxillary sinuses was cut. This traced transverse line, which occupied each sinus from its posterior wall to the anterior wall was divided into 3 parts: anterior, middle, and posterior by sketching perpendicular straight lines to the aforementioned standard (Figure 1 and 2).

The incidence and location of the septa in the panoramic radiographs and the CT scans were registered and categorized as primary septum (if it appeared from a maxillary tooth root) or secondary septum (if it appeared without teeth in the area). The lengths of the septa in the CT scans were measured and given consideration only if they measured more than 2.5 mm.8 

In the panoramic radiographs, when measuring the length of the alveolar processes, a halfway point was marked on a transversal straight line that occupied a region, and a perpendicular straight line was traced to this point until the alveolar process was cut. In this way, the length of the alveolar process was measured in the 3 areas of the sinus (anterior, middle, and posterior) (Figure 1) and was compared with the CT scans (Figure 2). The CT scans were carried out in the same manner using 3 panoramic cuts (a central cut corresponding with the M3 cut of the Dentascan, another more vestibular being B4 on the Dentascan, and another more palatal corresponding to the L2 cut on the Dentascan). The arithmetic mean of the lengths obtained in each cut of the CT scans (Figure 2) was then calculated and compared with those of the panoramic radiographs (Figure 1).

Among the 60 maxillary sinuses analysed, 14 were from group 1 (totally edentulous) and 46 were from group 2 (partially edentulous). According to the panoramic radiographs of 10 patients, 5 had 1 septum in 1 sinus, 3 had 1 bilateral septum, and 2 had 2 unilateral septa, thus making a total of 15 septa in 13 maxillary sinuses (21.7%). In the other 20 patients no septa were found. The septa were classified as to type and location: 3 were primary and 12 secondary. The septa were located as follows: 6 anterior, 7 middle, and 2 posterior.

In the CT scans of 11 patients, 5 had a septum in 1 sinus, 4 had a bilateral septum, and 2 had 2 unilateral septa, making a total of 17 septa in 15 maxillary sinuses (25%). The septa were classified as to types and locations: 3 were primary and the remaining 14 were secondary. The septa were located as follows: 2 anterior, 11 middle, and 4 posterior. Their heights were measured only in CT scans and registered between 2.5 and 6 mm long.

Comparing the incidence of septa observed in panoramic radiographs and CT scans 2 false-negatives occurred (11.8% of possibility of not locating septa).

The average values of the length of the subantral alveolar process in the anterior, middle, and posterior regions may be observed in Table 1. The fewest differences between the 2 techniques were observed in the middle regions of the sinuses. The difference in millimeters when calculating the height of the alveolar process between both techniques varied between 0.5 and 2.2 mm and varied the least in the middle region. The average length has a statistically significant value for these measurements.

The panoramic radiographs and the CT scans are the most frequently used radiologic methods for planning dental implants.5^,9^,10 The incidence of antral septa in the CT scan was found to be between 13% and 36% (Table 2).1^,3^,5,^–10 When comparing its incidence in both radiologic techniques, several authors7,^–10 have observed false-negatives in the panoramic radiographs. Some authors7 have had results with 50% false-negatives. Therefore, the presence of septa is frequent and is important in planning sinus floor elevations. These studies indicate that CT scans are more reliable than panoramic radiographs in the preoperative analysis.

As for the location of the septa, several studies5^,6 have observed them to be in the anterior regions, another1 found greater numbers in the posterior regions, and others8 (which agreed with our studies) noted a greater incidence in the middle regions.

Regarding the length of the septa, several authors obtained an average interval of 6.4 to 12.7 mm,1^,5^,6^,10 but Velásquez-Plata et al,8 unlike the others, reported lower average heights, varying from 2.7 to 8.3 mm. These results were similar to those found in our study.

In the study of the heights of the alveolar processes, Bolin and Eliasson,11 in 683 partially and totally edentulous regions, compared the distance from the floor of the sinus to the edge of the alveolar process and compared the panoramic radiographs and the CT scans. As in other studies there was a greater distortion in the panoramic radiographs than in the CT scans. However, the distortion was found to be less in the middle regions using the panoramic radiographs. Other authors11^,12 agree with our findings in observing fewer variations in the subantral lengths of the middle region. These values are closer to those obtained using the CT scans of the subantral middle region, and these lengths were independent of the type of edentulism of each patient.