The placement of implants in the posterior maxillary area is considered a reliable procedure, offering recognized rehabilitative advantages. The aim of this study was to evaluate the performance of dental implants placed in the sinus floor augmented with a block autograft by comparing the outcomes over 5 years with those of dental implants positioned in non-augmented bone. This retrospective cohort study included 16 patients who had undergone prosthetic rehabilitation supported by dental implants between 2000 and 2006. One implant per patient was included and assigned to one of two predictor groups: grafted versus ungrafted maxillary sinus. Changes in marginal bone level (MBL) and apical bone level (ABL) over time, at 1, 3, and 5 years, were the primary outcome variables. Appropriate pair-wise comparison tests were performed. No significant differences were seen with regard to ABLs and among times between the grafted group (nine implants) and the ungrafted group (seven implants). Significant marginal bone resorption was found over time, primarily at the buccal aspect, in both study groups. The bone surrounding the apex of dental implants appeared stable after sinus augmentation in the grafted area. The behaviour of the two groups with regard to loss of MBLs over time was very similar.
Patients who are affected by severe maxillary atrophy require reinforcement of the supporting bone if a fixed prosthetic rehabilitation through osseointegrated dental implantation is needed. When posterior maxillary areas are involved in the rehabilitation, and a modification of the crown/root ratio via augmentation of the alveolar crest is not helpful, dental implants can be successfully placed beneath the existing maxillary sinus if pneumatization has not occurred. However, if the vertical dimension of the residual bone is insufficient for implant placement, a maxillary sinus floor lifting procedure is unavoidable.
Different materials have been employed for this purpose, both non-autogenous, including synthetic bone graft, allograft, and xenograft, and autogenous (the material that still most satisfies the requirements of an ideal graft), with linear and volumetric analyses showing that all these materials are resorbed over time.
Various authors have described peri-implant bone resorption, which has been the subject of several reviews, although it should be noted that data from studies that have not excluded advanced surgery (for example a sinus lift operation), or which have been related to studies exclusively limited to maxillary implants, are limited in number. With regard to studies describing the apical modification of materials used to increase the bone volume beneath the maxillary sinus floor, a pneumatization phenomenon has been reported, with particulate autogenous bone appearing to be more sensitive to this event. Nevertheless, some authors have asserted that inorganic additives or xenogeneic materials used as bone substitutes have demonstrated no positive influence on the resorption rate, with an observed reduction in bone height during the observation period. Only a few of them have reported a progressive bulging of the apex of dental implants into the sinus cavity. Moreover, a recent review paper suggested that the risk of bone resorption and sinus pneumatization could be reduced with the use of a mixture of particulated autogenous bone and bovine bone materials or alloplastic materials, such as bovine bone mineral or hydroxyapatite. The short-term resorption of the materials surrounding dental implants inserted into autogenous and allogeneic grafts appears to be restricted to a percentage of <25%, and this is attested to the maintenance of the apical cortical plate of the block and to the absence of ‘bulging’ of the implant apex in the sinus.
The purpose of this study was to analyze the long-term changes in autogenous bone placed into the maxillary sinus in block form by comparing them to those of ungrafted areas. The investigators hypothesized that the compact, dense, and thick nature of the corticocancellous graft might guarantee bone preservation when the autogenous bone is inlay grafted into the maxillary sinus.
The specific aims of the study were: (1) to compare apical and marginal bone resorption around dental implants placed in the sinus floor augmented with a block autograft with that of dental implants positioned in non-augmented bone beneath the maxillary sinus; (2) to compare bone resorption for each of the groups (grafted and ungrafted) at three follow-up times (1, 3, and 5 years); and (3) to investigate the survival of the dental implants placed.
Materials and methods
In order to investigate the present assumption, the investigators designed and implemented a retrospective cohort study. The patient population comprised all subjects requiring fixed prosthesis rehabilitation who were treated with endosseous dental implants in the posterior maxilla between January 2000 and December 2006. All patients were treated at the study hospital.
Patients were included as study subjects if they met the following inclusion criteria: (1) at least one dental implant placed in the posterior maxilla either in ungrafted bone, beneath the sinus floor, or in bone augmented with an autologous bone block sinus grafting procedure; (2) presence of a complete set of computed tomography (CT) scans (acquired at 1 year (T 1 ), 3 years (T 2 ), and 5 years (T 3 ) after dental implant insertion).
Patients were excluded as study subjects if they had undergone adjunctive surgical procedures performed on the alveolar crest in the same posterior maxilla, or bone resection and irradiation as a result of oncologic treatment; they were also excluded if they had been subjected to a pharmacological treatment affecting bone turnover (such as bisphosphonate drugs).
Surviving dental implants placed in the posterior maxillary area were divided into two groups: (1) those in an ungrafted area, comprising dental implants placed in native bone beneath sinuses that had not undergone surgical augmentation, and (2) those inserted in augmented bone following a sinus lift and autogenous block graft procedure. One implant per patient was selected randomly using a computerized random allocation process. The present retrospective cohort study was approved by the ethics committee of the study institution.
Variables were divided into those for sample description and those that defined the sample composition, the latter being the primary predictor variables that were used to classify patients into a finite number of subgroups. The radiological measurements of bone level were the numerical input variables obtained directly by the clinicians and related to the outcome. The outcome variables (changes in bone levels) were obtained from the input variables.
Variables for sample description
Sample description variables were age, gender (male or female), smoking habit (smoker or non-smoker), location, length, and diameter of the dental implant placed, and the thickness of the residual sinus floor measured before surgery at the site planned for the dental implant placement. The number of adjunctive simultaneous bone augmentation procedures in areas other than the enrolled maxillary posterior area (contralateral or anterior site), if such procedures had been required, was also recorded.
Primary predictor variables
The primary predictor variable was the type of bone in which the implant was placed, i.e., an ungrafted area, or a grafted area.
The input variables obtained by the clinicians were the following: (1) the distances between the apex (the axis origin) and bone level of the sinus floor measured for all four aspects: buccal, palatal, mesial, and distal (apical bone levels (ABLs)). When the position of the new sinus floor was coronal to the implant apex, the ABL vector was negative, otherwise positive. ABL 1 , ABL 2 , and ABL 3 were the ABLs at 1, 3, and 5 years, respectively. (2) The marginal bone level (MBL) values were obtained from the arithmetical difference between the crestal bone height (measured with the apex of each dental implant as the axis origin) and the implant length for the same four aspects. MBL 1 , MBL 2 , and MBL 3 were the MBLs at 1, 3, and 5 years, respectively.
The outcome variables of interest, describing hard tissue modifications from the first to the third year of follow-up, and from the third to the fifth year of follow-up, were negative regarding resorption, and were summarized as follows: (1) changes in the apical bone levels (ΔABL) evaluated for all four aspects, by subtracting the value of the preceding follow-up from the next one; for site n, nΔABL X | X +1 = nABL X +1 − nABL X , with X = 1, 2, obtaining nΔABL 1|2 and nΔABL 2|3 , respectively, with n as mesial, distal, buccal, or palatal. The overall changes in apical bone levels (ΔABL 1|3 ) were obtained by adding the two partials. (2) Changes in the marginal bone levels (ΔMBL) were evaluated for all four aspects, in the same way as ΔABL regarding the intervals between the follow-up times. Two values of change in MBLs were obtained: nΔMBL 1|2 and nΔMBL 2|3 , with n as mesial, distal, buccal, or palatal. Overall changes in marginal bone levels (ΔMBL 1|3 ) were obtained by adding the two partials.
Data collection methods and summary of operative methods
Patient suitability for dental implant placement in the posterior maxillary area had been assessed by preoperative CT scan. CT scan follow-up (High-Speed Double Detector CT Scanner; GE Healthcare, Milwaukee, WI, USA) was then performed after implant insertion. A surviving implant was defined as being non-mobile, free from peri-implant radiolucency, infection, and neurological disorder, and without associated pain, either spontaneous or on the application of pressure.
The dental implant apex was identified both in cross-section and in panoramic radiograph view, and was used as the axis origin; distances between the apex and both the MBLs and ABLs were acquired at a distance of 0.5 mm buccally, palatally, mesially, and distally from the peri-implant outline. Numerical measurements were made to the nearest tenth of a millimetre by a single investigator who was not involved in the treatment of the patients.
Surgical procedures, implant insertion, and prosthetic loading and timing were performed as described in previous papers. In brief, the iliac crest area was used as an extraoral harvesting site and grafts were obtained in accordance with the technique described by Grillon et al., using a cutaneous approach via elective lines of incision. The harvested bone was then moulded and used as a single block.
Approaching the recipient site, a horizontal midcrestal incision and two vertical releasing incisions were performed to expose the lateral wall of the maxillary sinus. Osteotomy lines were then drawn in order to provide adequate access for the positioning of the planned bone block graft. The moulded corticocancellous bone was placed and secured to the denuded antral floor by fixation screws, positioned with a lag screw technique. The bone grafts were then left to heal for 3–4 months after reconstructive surgery, and delayed titanium dental implants (root-form, external-hex, sand-blasted, and acid-treated rough-surfaced screws), all of the same brand, were placed. No ‘platform-switched’ restoration was applied.
Normal distribution of the samples was verified by the Lilliefors test. For each of the radiological measurements and outcome variables, the influences among primary predictor variables and those for sample description were analyzed by multi-factor analysis of variance (ANOVA) for multiple comparison tests with Tukey’s honestly significant difference criterion, and a P -value for two-variable interactions was computed. Influences between and within subject variables (primary predictor variables and times) were tested using a repeated measures mixed two-way ANOVA test. Then, for a more conservative analysis, significance was assessed for each pair-wise comparison using the Wilcoxon signed rank test and the Wilcoxon two-sided rank sum test for matched data (comparisons of times) and unmatched data (primary predictor group comparisons), respectively. Spearman correlation ( r s ) was used to assess the strength of the bivariate association between radiological and outcome variables and each of the variables describing the sample, separately for each primary predictor variable.
An estimation of sample size was determined through a power analysis (power of 90%), based on measurements of central tendency and dispersion. All measurements in the text and tables are given as the mean and standard deviation (SD). Statistical analyses were performed using Statistics Toolbox for MatLab 7.0.1 (The MathWorks Inc., Natick, MA, USA) and the level of significance was set at 0.05 for all analyses.
Sixteen patients, all partially edentulous, were enrolled; five were females, three of whom were smokers, and 11 were males, one of whom was a smoker. Patient demographic characteristics and implant data (the residual bone height at the site at which the implant was placed and the physical specifications (diameter and length) of the dental implant inserted) are given in Table 1 . A total of 37 dental implants were placed in the posterior areas beneath the maxillary sinus. Sixteen of these implants were in native bone, of which seven were included in the numerical computation. Nine implants were included from among those placed in a grafted sinus. The total number of inlay–onlay blocks grafted in the patients is also given in Table 1 : two patients with an enrolled implant in an ungrafted sinus also underwent bone augmentation in other maxillary areas; three patients with an enrolled implant placed in a grafted maxillary sinus underwent no further simultaneous bone augmentation procedures. In total, eight patients were subjected to adjunctive simultaneous surgical augmentation procedures (adjunctive grafts, from one to three blocks).
|Patient||Age, years||Gender||Smoking habit||Implant in bone||Residual bone height, mm||Adjunctive simultaneous bone augmentation procedures, n||Position of enrolled implants||Diameter, mm||Implant length, mm|
Apical peri-implant radiographic bone remodelling
ABLs and the respective changes (Δ) in values for both ungrafted and grafted sinuses at all follow-up times are shown in Figs 1 and 2 . Results of the statistical analysis are given in Table 2 ; pair-wise comparisons were made and the estimated number in the sample size, obtained through a power analysis, resulted in values not larger than 50 with matched P -values.
|Variables||Time or type of bone||Time or type of bone A||Time or type of bone B||A||B||ANOVA test: (1) predictor, (2) time, (3) predictor × time||Pair comparisons (number of implants)||P -value, Wilcoxon test a||Estimated effect size b|
|Apical measurements and outcomes|
|b ABL||T 1||Grafted||Native||1.1 (1.6)||−0.8 (5.2)||(1) 0.1906||9 vs. 7||0.2944 d||42|
|b ABL||T 2||Grafted||Native||1.1 (2.9)||−0.5 (2.9)||(2) 0.8403||9 vs. 7||0.2845 d||36|
|b ABL||T 3||Grafted||Native||1.8 (2.5)||−0.7 (4.5)||(3) 0.8083||9 vs. 7||0.1749 d||25|
|p ABL||Native||T 1||T 2||3.0 (2.5)||1.7 (2.9)||(1) 0.4365||7 vs. 7||0.2039 c||49|
|p ABL||Native||T 1||T 3||3.0 (2.5)||1.0 (3.0)||(2) 0.1709||7 vs. 7||0.1908 c||23|
|p ABL||T 1||Grafted||Native||1.4 (2.3)||3.0 (2.5)||(3) 0.5477||9 vs. 7||0.2183 d||28|
|m ABL||T 1||Grafted||Native||1.5 (2.1)||2.9 (2.1)||(1) 0.5422||9 vs. 7||0.2362 d||30|
|d ABL||T 2||Grafted||Native||0.5 (2.2)||2.0 (2.4)||(1) 0.7782||9 vs. 7||0.2149 d||28|
|p ΔABL||T 1 → T 3||Grafted||Native||−0.6 (1.3)||−2.0 (3.6)||(1) 0.2910||9 vs. 7||0.4079 d||41|
|m ΔABL||T 1 → T 3||Grafted||Native||0.7 (2.4)||−0.7 (3.0)||(1) 0.3252||9 vs. 7||0.2991 d||44|
|d ΔABL||T 2 → T 3||Grafted||Native||0.8 (4.5)||−1.1 (3.2)||(1) 0.4516||9 vs. 7||0.4698 d||44|
|Marginal measurements and outcomes|
|b MBL||Grafted||T 1||T 2||−2.4 (1.0)||−3.2 (1.9)||(1) 0.0520||9 vs. 9||0.3819 c||43|
|b MBL||Grafted||T 2||T 3||−3.2 (1.9)||−4.3 (1.8)||(2) 0.0020 *||9 vs. 9||0.0453 c, *||31|
|b MBL||Grafted||T 1||T 3||−2.4 (1.0)||−4.3 (1.8)||(3) 0.9086||9 vs. 9||0.0178 c, *||9|
|b MBL||Native||T 1||T 2||−2.7 (2.3)||−3.6 (1.6)||7 vs. 7||0.3762 c||50|
|b MBL||Native||T 2||T 3||−3.6 (1.6)||−5.0 (2.5)||7 vs. 7||0.0856 c||26|
|b MBL||Native||T 1||T 3||−2.7 (2.3)||−5.0 (2.5)||7 vs. 7||0.0414 c, *||14|
|p MBL||Grafted||T 1||T 2||−1.4 (1.0)||−2.6 (1.4)||(1) 0.9769||9 vs. 9||0.0255 c, *||13|
|p MBL||Grafted||T 1||T 3||−1.4 (1.0)||−2.3 (0.9)||(2) 0.0041 *||9 vs. 9||0.0001 c, *||14|
|p MBL||Native||T 1||T 2||−1.5 (0.5)||−2.4 (1.6)||(3) 0.9119||7 vs. 7||0.2140 c||20|
|p MBL||Native||T 1||T 3||−1.5 (0.5)||−2.4 (0.9)||7 vs. 7||0.0234 c, *||9|
|m MBL||Grafted||T 1||T 2||−1.5 (0.6)||−2.2 (1.4)||(1) 0.3194||9 vs. 9||0.1653 c||26|
|m MBL||Grafted||T 1||T 3||−1.5 (0.6)||−2.2 (0.6)||(2) 0.0089 *||9 vs. 9||0.0130 c, *||11|
|m MBL||Native||T 1||T 2||−1.7 (1.1)||−2.7 (0.9)||(3) 0.8423||7 vs. 7||0.0639 c||14|
|m MBL||Native||T 1||T 3||−1.7 (1.1)||−2.6 (1.1)||7 vs. 7||0.0638 c||20|
|m MBL||T 3||Grafted||Native||−2.2 (0.6)||−2.6 (1.1)||9 vs. 7||0.3294 d||47|
|d MBL||Grafted||T 1||T 2||−1.5 (0.7)||−2.1 (1.0)||(1) 0.1120||9 vs. 9||0.1568 c||28|
|d MBL||Grafted||T 1||T 3||−1.5 (0.7)||−2.3 (0.8)||(2) 0.0529||9 vs. 9||0.0398 c, *||15|
|d MBL||Native||T 2||T 3||−2.2 (1.5)||−2.9 (0.9)||(3) 0.9809||7 vs. 7||0.3240 c||44|
|d MBL||T 1||Grafted||Native||−1.5 (0.7)||−2.2 (1.5)||9 vs. 7||0.2380 d||33|
|d MBL||T 2||Grafted||Native||−2.1 (1.0)||−2.8 (0.9)||9 vs. 7||0.2228 d||28|
|d MBL||T 3||Grafted||Native||−2.3 (0.8)||−2.9 (0.9)||9 vs. 7||0.1892 d||25|
|p ΔMBL||Grafted||T 1 → T 2||T 2 → T 3||−1.2 (1.3)||0.3 (1.5)||(1) 0.9984||9 vs. 9||0.3008 c||13|
|p ΔMBL||Native||T 1 → T 2||T 2 → T 3||−1.0 (1.8)||0.0 (1.5)||(2) 0.1164||7 vs. 7||0.2969 c||33|
|m ΔMBL||Grafted||T 1 → T 2||T 2 → T 3||−0.7 (1.4)||0.1 (1.3)||(1) 0.5931||9 vs. 9||0.4258 c||36|
|m ΔMBL||Native||T 1 → T 2||T 2 → T 3||−1.0 (1.19)||0.1 (0.9)||(2) 0.1285||7 vs. 7||0.2188 c||12|