Abstract
The aim of this study was to establish an objective method for quantitative evaluation of bone volume change after sinus augmentation. 11 sinuses in 9 patients were evaluated by computed tomography images taken before treatment (T0), and 3 months (T1) and at least 1 year (T2) after sinus augmentation. Based on the 3D digital subtraction technique, augmented bone images were extracted and bone volumes were calculated from voxel numbers of the extracted images. The mean augmented bone volumes at T1 and T2 were 2.46 cm 3 and 1.85 cm 3 , respectively. These bone volume changes were statistically significant and the mean bone volume change ± SE was −24.8% ± 6.1%. Loss of augmented bone was observed in all except one of the patients. The correlation coefficient between bone volume change and elapsed time was −0.64, which was statistically significant and indicated that bone resorption progressed with elapse of time after sinus augmentation. The authors’ method of analysis enabled visualization of augmented bone and objective assessment of bone volume change. Within the limited number of cases, the present investigation demonstrated a significant decrease in augmented bone volume between 3 and 23 months after surgery.
Enlarged maxillary sinuses and lack of vertical bone in the posterior maxilla frequently preclude proper implant placement in the region. Internal augmentation of the maxillary sinus floor was first described by Boyne and James in 1980 and has become widely accepted as one of the most predictable procedures for improving bone volume before implant placement.
Various grafting materials such as autografts, allografts, and synthetic bone grafts have been used for sinus augmentation, but it has been reported that such bone graft materials were resorbed over time. Volume and long-term outcome of bone grafts in the maxillary sinus are of great interest in clinical research, and several techniques for assessing augmented bone changes have been described. Panoramic radiography allows estimation of the vertical dimension of grafts but does not provide information about volume and three-dimensional (3D) changes. Computed tomography (CT) offers a reliable technique for 3D visualization of regenerated bone formation after sinus floor augmentation, and volume changes of augmented bone in the maxillary sinus were evaluated in some studies by using CT images, though the grafted area was marked manually in each CT slice.
In the authors’ method using CT images, augmented bone after sinus floor augmentation was identified by the 3D digital subtraction technique using longitudinal CT data before and after surgery. In this pilot study, the volume and structural changes of the augmented bone were evaluated over time retrospectively in patients treated by sinus floor augmentation with autogenous bone grafts.
Materials and methods
The subjects consisted of 9 patients (3 males and 6 females, aged 35–64 years) in whom maxillary sinus augmentation procedures were performed in Niigata University Medical and Dental Hospital ( Table 1 ). In all cases, the preoperative alveolar bone height of the posterior maxilla was <5 mm on CT images and a surgical approach for sinus floor augmentation was necessary for the implant insertion to succeed. Two patients underwent bilateral sinus augmentation and seven patients underwent unilateral sinus augmentation. The presented evaluation is based on a total of 11 augmentation sites. The study protocol was approved by the Ethics Committee of Niigata University and informed consent was obtained from the patients before surgery.
| Patient No. | Gender | Age | Side | Donor site | Period between T1 and T2 (months) |
|---|---|---|---|---|---|
| 1 | F | 57 | Both | Iliac crest | 10 |
| 2 | M | 55 | Both | Iliac crest | 18 |
| 3 | F | 50 | Left | Mandibular ramus | 39 |
| 4 | F | 35 | Right | Mandibular ramus | 20 |
| 5 | M | 64 | Right | Mandibular ramus | 9 |
| 6 | M | 55 | Right | Mandibular ramus | 44 |
| 7 | F | 50 | Left | Mandibular ramus | 16 |
| 8 | F | 59 | Left | Mandibular ramus | 32 |
| 9 | F | 37 | Left | Mandibular ramus | 32 |
Sinus floor augmentation was performed with the lateral window technique described by Boyne and James. In brief, the posterior part of the maxilla was exposed by a crestal incision and elevation of a mucoperiosteal flap, and osteotomy was performed at the lateral aspect of the sinus wall. After carefully lifting the sinus mucosa, autogenous bone harvested from the iliac crest (4 sinuses in 2 patients) or mandibular ramus (7 sinuses in 7 patients) was crushed by a bone mill, and the sinus floor was filled with the crushed bone to augment the bone height without any artificial materials. The choice of donor site for autogenous bone depended on the volume needed. In two cases with bilateral sinus floor augmentation, a relatively large amount of harvested bone from the iliac crest was grafted, and harvested bone from the mandibular ramus was used for unilateral sinus floor augmentation in the other cases.
Twenty-four solid titanium screw-type implants including Straumann, Brånemark and Ankylos implants were placed into the premolar or molar regions of maxillary alveolar bone several months (range 131–220 days) after sinus floor augmentation. All implants have been functioning for at least 16 months after implant placement without loss of implant, peri-implantitis and/or pain.
CT scans of the maxillary sinus region were performed before treatment (T0) and at 3 months (T1) and at least 1 year (T2) after sinus augmentation. The mean period between T1 and T2 was 23 months (range 9–44 months) ( Table 1 ). All scans were carried out using single detectable CT (Xvigor, Toshiba, Tokyo, Japan) or 64 multi-detectable CT (Aquillon, Toshiba, Tokyo, Japan) at Niigata University Medical and Dental Hospital. CT data were acquired by taking a 1.0 mm slice thickness, 0.5 mm slice interval and 1.5 helical pitch (Xvigor) or 0.5 mm slice thickness, 0.3 mm slice interval and 21 helical pitch (Aquillon), using 120 kVp and 100 mA, respectively.
Volumetric analysis
Volumetric measurement of the augmented bone was obtained from DICOM data of CT scans using Real INTAGE (Cybernet Systems Co., Ltd, Tokyo, Japan) ( Fig. 1 ). Initially, DICOM data were transformed into isotropic voxel size and the bone was identified if voxel data were over 50 Hounsfield units (HU). 3D images of the maxilla were constructed from DICOM data at T0, T1, and T2 and the coordinates of these images were adjusted and unified by superposition of four anatomical landmarks: anterior nasal spine (ANS), posterior nasal spine (PNS), and the lowest points of the pterygoid process on both sides. These images were cut to the same desired region of interest (ROI) involving the maxillary sinus. 27 CT images in 9 patients were analysed and 33 ROI images of 11 sinuses were obtained.
Using the 3D digital subtraction technique, the ROI images at T0 were logically subtracted from those at T1 and T2 and the augmented bone images at T1 and T2 were extracted three-dimensionally. Colour mapping of the images according to the classification described by Misch was applied to evaluate bone quality. In this classification, bone density was determined using Hounsfield units of CT data as follows: D1 >1250 HU; D2 850–1250 HU; D3 350–850 HU; D4 150–350 HU; and D5 50–150 HU.
The augmented bone volumes at T1 and T2 were calculated from the voxel number of the extracted images. The measurements were repeated three times by the same researcher and the mean values were employed as individual data. The bone volume changes were calculated from the measured bone volumes at T1 and T2.
Statistical analysis
The significance of statistical differences between the augmented bone volumes of 11 sinuses at T1 and T2 was assessed by the paired t -test because normal distributions of bone volumes were confirmed by the Shapiro–Wilks test. The relation between volume change of the augmented bone and elapsed time was evaluated with Pearson’s correlation coefficient. All statistical analysis was carried out with the IBM SPSS Statistics Base 18 for Windows (SPSS, an IBM Company, Tokyo, Japan). A probability of <0.05 was regarded as statistically significant.
Results
Augmented bone images
The images of augmented bone were clearly displayed and observed three-dimensionally, and comparison of identical cross-section images at T1 and T2 was useful for the alteration of components in the augmented bone ( Fig. 2 ). Some superficial resorption on the superior region of the augmented bone was observed in all cases at T2, though bone enlargement was observed in patient No. 2 because an area with lower density under 50 HU observed at T1 changed to dense bone at T2. In almost all cases, there was obvious qualitative improvement of augmented bone surfaces at T2 as indicated by colour changes compared with those at T1. No implant was exposed on the augmented bone surface.
