Virtual Implant Rehabilitation Of The Severely Atrophic Maxilla: A Radiological Study

Background: Advanced maxillary atrophy is often observed and implant placement could become difficult. Nevertheless, a volumetric evaluation using a proper diagnostic software could facilitate the implant planning. The purpose of the present study is to suggest the potential application of the maxillary retro-canine area as the designated place for virtual tilted implants also in severely atrophic maxillae. Methods: A sample of CBCT images from the Department of Dentistry (IRCCS San Raffaele, Milan) was evaluated. After a 3D anatomical evaluation of the maxilla tilted implants has been virtually positioned in the retro-canine regions. All the implants were inserted with the same procedure at 30° and 45° degrees of tilting. The length, the palatal angulation and the diameter of each implant were identified. Results: 220 tilted implants were placed. An average implant measurement of 13.508 mm of length and 3.42 mm of diameter were calculated. Also, an average buccal-palatal angulation of 6° was identified. After the statistical analysis implant length was found significantly higher at 45° degrees than at 30° degree (<0.0001). Conclusions: A considerable amount of patients show a significant degree of bone atrophy. The implant-supported treatment plan can rely on the three dimensional imaging of the residual bone as a guiding tool to establish the most effective implant position for each specific case. In this study it is founded that an implant could be longer if its mesio-distal angulation is more accentuated. In addition, it is possible to virtually insert tilted implants also in severely atrophic maxilla. This could help clinicians to consider the retro-canine area a viable place to insert longer tilted implant.

requirements for dental implants osseointegration is the presence of a sufficient amount of basal bone 7 . Unfortunately, totally or partially edentulous maxillae often show a significant degree of sinus pneumatization and/or alveolar bone atrophy 8 . This lack of bone may complicate the implants placement and could influence the final results 9 . For these reasons, several aspects should be considered before the implant surgery and a 3D volume bone evaluation is essential to plan a proper implant rehabilitation 10 .
The Cone Beam Computed Tomography (CBCT) produces three-dimensional reconstructions of maxillary and mandibular anatomical structures using a single scan and offering multiple views with low radiations. A flat detector makes the capture. The X-ray diffusion is cone shaped. The CBCT scans permits to gain a better understanding of the jaws' morphology and to evaluate the volume of remaining bone in any given site. Furthermore, the scans allow interactive planning using 3D simulation software. To date, according to the American Academy of Oral and Maxillofacial Radiology, CBCT should be considered as the method of choice for the three dimensional evaluation of the maxillary bone to plan an implant treatment 11 . In particular, volumetric data acquired by CBCT showed a high accuracy of the measurements with a relative error below 1% compared with the same measurements took in vivo as demonstrated by Veyre-Goulet et al 12 . Many surgical techniques are suitable to place and load implants in edentulous atrophic maxilla. Treatment plans may include bone grafting techniques to reconstruct the lost bone volume 13 . For example procedures such as distracting osteogenesis, Le Fort osteotomy with inlay grafting, onlay bone grafts, maxillary sinus floor elevation or guided bone regeneration have been used to re-establish bone bulk before implant surgery [14][15][16][17] . However, these methods are all characterized by high morbidity and long duration of the therapy [18][19] . In addition, patient satisfaction with graftless solutions for implant rehabilitation of edentulous jaws is generally high 20 . Widmark et al. 21 found that implants inserted in native bone had a higher success rate than implants placed into grafted bone. The trans-sinus and zygomatic implants are viable alternatives for treatment of the severely atrophic maxilla, nevertheless this procedures are operator-dependent 22 .
In this study, tilted implants insertion is simulated on CBCT scans of atrophic upper jaws in order to calculate the average angulations and measures of the fixtures.
The aim of the present paper is to suggest the potential application of the maxillary retro-canine region as the designated place for virtual "iuxtameatal" implants. "Iuxtameatal" means that the apex of the implants is surrounded by the cortical bone near the nasal inferior meatus. This virtual procedure could help surgeons finding the most effective implant measures and position in order to get more implant stability and considering a tilted implant rehabilitation without the need of sinus lifts or invasive bone grafting operations in patients presenting a severe maxillary atrophy.

Study Population
Patients were retrospectively selected among those referred for a Cone Beam Computerized Patients were not included if they presented one of the following exclusion criteria: Radiolucent or radiopaque images in the mid-maxilla area Implant or impacted tooth in the mid-maxilla area The "mid-maxilla" region is an anatomical district that belongs to the maxillary bone. It is extended from the lateral nasal cavity's wall until the medial wall of the maxillary sinus cavity, including also the residual alveolar process below the sinus floor (Fig. 1). The mid-maxilla could be also identified with the retro-canine bone triangle [23][24] .
Each patient provided a written consent before undergoing the CBCT scan. No one refused to sign the informed consent.
All CBCT scans were acquired with a Field of View of 12 x 8 cm, at 90kV 10 mA 16 s and 0,2 mm Voxel size with a NewTom VGi evo Cone Beam 3D Imaging Device (Cefla SC, Imola Italy). The default position and orientation of the orthogonal sectional planes relative to the jaws were consistent in all the CBCT datasets of each patient. To guarantee a stable head position, all CBCT scans were checked and re-orientated to place the scan view parallel to the Camper's plane.

Virtual Iuxtameatal Implant Positioning
The residual bone of each scan was analyzed using the panorex imagine construction (Fig. 2).
According to the software procedures the panorex image was obtained drawing a panoramic curve on the axial view of the maxillary segment above the residual alveolar crest. The cross sectional views perpendicular to the panoramic curve were automatically elaborated by the software.
For each patient, two iuxtameatal cone shape implants were virtually placed in tilted position in each side, at 30° and at 45° degrees of mesio-distally angulation (Fig. 3). The majority of implant companies produces abutments which can correct an angle of 30° maximum, but the prosthesis meso-structure can correct additional 15° of angulation with its conic component, so that the maximum of angulation could be 45°. All tilted implants were placed in analogous way, according to the All-on-four procedure described by Malò et al. 25 trying to maximally exploit all the bone volume offered with at least 2mm of bone all around the fixtures (Fig. 4).

Outcome measures
Once the iuxtameatal implants were positioned in the mid-maxilla, the length and the diameter of the implants were measured. The buccal-palatal angulation of each implant were analyzed too in the cross sectional views. A negative angle indicated that the implant insertion went from the palatal side to the vestibular side of the residual alveolar crest. With the dedicated software also the bone density around the implants' apex was measured in gray-scale. All measurements were performed using a digital ruler at 0.1 mm increments by a single trained examiner (A.M).

Statistics
All statistical analyses were performed with a specific software (R, R Core Team, Foundation for Statistical Computing, Vienna, Austria). To evaluate the effect of the tilting degree (30° vs. 45°d egrees) on the implant length, a linear mixed-effects model was estimated. In particular, the Linear and Nonlinear Mixed Effects models package was used to estimate the Linear Mixed Effects model The modelling approach here applied allows to properly account for repeated measure data and for unobserved heterogeneity among patients. Actually, along with fixed effects, the model allows to specify in the model random components. An initial complete model was estimated including position and tilting degree as fixed effects along with their interaction. Subject-specific random effect was specified. Hence, a random intercept model was considered. Assumptions for the correct application of the model were checked. A backward stepwise procedure was applied to select a more parsimonious model.
In all the analyses, the significance threshold was set at 0.05.

Results
CBCT scans of 59 subjects (28 male and 31 females; mean age: 64.5 ± 8.2 years) were included in this study according to the inclusion criteria. Only two scans showed a very severe maxillary atrophy in both right and left maxillae. Among the other 57 scans, one of them presented the right maxilla completely reabsorbed and other three presented the left maxilla too narrow to place any implant.
A total of 220 iuxtameatal implant insertions were simulated (see additional file 1). An average implant measurement of 13.508 mm of length and 3.42 mm of diameter were calculated (Tab. 1).
Also, an average anterior-posterior angulation of 6° were identified. In some simulations the implant axis was negative: the fixtures' most coronal point was palatal respect of the apex. The average bone density around implants was 570 gray-scale. After the statistical analysis implant length was found significantly higher at 45° degrees than at 30° degree (<0.0001). Thus, when considering tilting degree of the implant, significant effects of implant length were found (Fig. 5).

Discussion
Krekmanov et al. 26 and Aparicio et al. 27 presented the first papers in which a combination of tilted and axial implants was used in patients with severely reabsorbed posterior maxillae. The results indicate that the use of tilted implants is an effective and safe alternative to maxillary sinus floor augmentation or bone grafts procedures.
Peñarrocha-Oltra et al. 28 in 2013 wrote that tilted implants, both used alone and combined with axially placed implants and rehabilitated with different prosthetic options have high success rates, minimal complications and high patient satisfaction.
Additionally, Balleri et al. 29 presented a very good outcome with 20 fixed partial dentures supported by two implants, one tilted and one axial, in the retro-canine bone triangle. In a recent finite element study for two splinted implants, it appeared that tilting of the distal fixture does not stress the periimplant bone as compared with the mesial axial fixture 30 . It was also demonstrated that tilted posterior implants were mechanically more advantageous than distal cantilever units 31 . Finally, the study of Barnea 32 demonstrates no effect of implant angulation on peri-implant bone loss in the posterior maxilla. Nunes et al. 33  Owing to mechanical and anatomic difficulties, implant treatment in the atrophic maxilla represents a challenge.
The short and ultra-short implants in the posterior areas could be an alternative, but the trabecular compartment of the canine/premolar region offered higher quality when compared with the posterior maxilla. In addition, the basal bone showed a higher density than the alveolar bone. Corroborating the statement, Gonda et al. 36 showed that premolar areas offer the most favourable scope in terms of bone height, width, angulation and quality.
In this paper the examiners were able to find enough bone to adequately distribute the virtual tilted implants in all cases except two. This study presents some limitation such as the retrospective nature of the present analysis and the virtual placement of the implants. Nevertheless, the statistical analysis demonstrates that an implant could be longer if its angulation is more accentuated. Hence, tilting implants would allow inserting the fixture apex in a high bone density area. Longer iuxtameatal implants in the mid-maxilla area consent more implant stability and, eventually, an immediate prosthetic load [37][38] .
Benefitting from the virtual plan previously developed, the implants will unlikely exceed the cortical lamina of the nasal lateral wall. In the eventuality that it happens, the associated risks may be epistaxis, implant displacement in the nasal cavity and rhinitis. Nonetheless, this statement has to be confirmed through future clinical studies.
These tilted implants do not compromise implant placement in the anterior maxilla because they have a marked angulation in the palatal sense. But in the ortho-panoramic radiography the correct vestibular-palatal angulation of the implants cannot be planned: only CBCT scans can suggest the ideal angulation with extremely high precision. With this paper, clinicians could become aware of the importance of the 3D anatomical view to evaluate the amount of cortical bone around the nasal cavity and could learn the importance of simulation software to virtually insert implants also in severely atrophic maxillae.
Among CBCT scans gray-scales vary widely due to different factors, such as the lack of grey level uniformity, the presence of artifacts, the effects of scatter and beam hardening 39 . On the other hand, different studies demonstrated how grey levels of CBCT can be used to derive Hounsfield units [40][41] .
The gray-scale outcomes reported in this study could suggest that the peri-implant bone density was greater than the average density of the cancellous bone. It indicates that the apex of the iuxtameatal fixtures is effectively inserted in the cortical bone of the walls of the nasal meatus. Availability of data and materials The dataset supporting the conclusions of this article is included within the article (and its additional file). All materials described in this manuscript, including all relevant raw data, will be freely available to any scientist wishing to use them for non-commercial purposes, without breaching participant confidentiality.

Competing interests
The authors declare that they have no actual or potential financial relationships with any companies whose products or services may be related to the subject matter of the article.

Funding
The authors state that this work has not received any external funding. All authors read and approved the final manuscript.

12.
Veyre-Goulet S, Fortin T, Thierry A. Accuracy of linear measurement provided by cone beam computed tomography to assess bone quantity in the posterior maxilla: a  Figure 1 The mid-maxilla region is shown in this fresh frozen ex vivo fixed specimen. The vestibular wall of the right sinus has been removed in order to expose the extension of the paranasal cavity. The residual alveolar bone around the canine region is the area of interest for tilted implant positioning.

Figure 2
Panorex image elaborated by the software from a maxillary CBCT Figure 3 In this 3D reconstruction the iuxtameatal implant was placed in the right maxilla with 45° of angulation, in the left side with 30° of angulation.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. Outcomes.xlsx