Immediate Maxillary Full-Arch Rehabilitation of Periodontal Patients with Terminal Dentition Using Tilted Implants and Bone Augmentation: A 5-Year Retrospective Cohort Study

Background: All-on-four protocols with tilted implants in the maxilla are used to rehabilitate the terminal dentition of the severe generalized periodontitis patients. Data on long-term biological complications are scarce. Methods: Eighty-four axial and forty-six tilted immediate implants have been placed in the extraction sockets of 23 patients according to a four–six implants protocol combined with ridge augmentation. Within 72 h, a provisional prosthesis was cemented to the implants; after 6 months, a cemented ceramic–metallic prosthesis was delivered. The patients were followed for up to 5 years. Results: The 5-year survival rate of the straight and tilted implants was 100% and 97.8, and the prosthetic one was 100%. Marginal bone loss (MBL) of the straight implants was 0.42 ± 0.67 and 0.59 ±1.01 mm on the mesial and distal sides; for the tilted, it was 0.37 ± 0.68 and 0.34 ±0.62 mm, and the differences were not statistically significant. Implant position, smoking, keratinized mucosal width, and cantilever did not affect MBL. Peri-implant mucositis involved 29.4% and 22.2% of the straight and tilted implants, respectively; peri-implantitis involved 5.8% and 4.4% of the straight and tilted implants, respectively, without statistical significance. Conclusions: This immediate loading protocol’s 5-year survival and success rates were high. No difference between the straight and tilted implants was found regarding survival, success rates, and MBL.


Introduction
The edentulous posterior maxilla is often associated with the pneumatization of the maxillary sinus and a reduced height of the alveolar ridge at sites to place implants [1]. Bone augmentation of the maxillary sinus is a standard and predictable surgical procedure in implant therapy; however, this procedure is associated with patient morbidity, possible surgical complications, high costs, and a long 6-12 month healing period [2]. Other treatment options are short implants, implant-supported fixed partial dentures with a distal cantilever, tilted implants, and implants placed in the zygoma or the tuberosity [3,4].
The use of tilted implants has gained popularity as a feasible option to treat the edentulous maxilla by means of implant-supported rehabilitation without requiring a grafting procedure [5]. With this technique, anterior implants are placed axially, and the most posterior ones are tilted to run parallel to the anterior wall of the maxillary sinus [4]. Implementing tilted implants (TIs) may offer several clinical advantages: (1) placing longer (2) Compare the MBL of tilted and axial immediately loaded implants placed concomitantly with bone augmentation; (3) Assess the impact on the MBL of other covariates, such as smoking, bruxism, cantilever, implant location, and the presence of keratinized mucosa.
We hypothesized that high survival and success rates for both the axial and tilted implants and similar marginal bone losses would be achieved.

Patients Selection
This survey covers all consecutive patients treated with the all-on-4-6 protocol by one periodontist (RK) over 7 years. All the participants meet the inclusion criteria described below. The treatment consisted of extracting the existing maxillary teeth and placing 4-6 immediately loaded implants in combination with bone augmentation. All patients signed an informed consent form to allow the retrospective use of their clinical data. The study was approved by the ethics committee of Tel Aviv University (no. 0000030-1).

1.
A diagnosis of generalized stage 4 grade C by a periodontist according to the new classification of periodontal diseases from 2018, based on clinical examination and radiographic evaluation [21]; 2.
Need for the extraction of all upper teeth, primarily due to periodontitis. Other reasons included extensive carious lesions and root/tooth fractures; 3.
Mandibular teeth or implants sustaining a fixed partial or complete arch restoration or consisting of natural dentition with a good or fair prognosis; 4.
Stable periodontal disease of the mandibular teeth; 7.
≥5 mm bone apical or palatal to the alveolar socket; 8.
Full or partial integrity of the socket walls after extraction; 10. Cone-beam tomographic (CBCT) examination was obtained before surgery.
Current or past metabolic bone disease; 3.
Medical treatment with bisphosphonates (both oral and IV); 4.
A history of radiotherapy or chemotherapy to the head and neck region; 5.
Lack of compliance.

Surgical Treatment
The presurgical evaluation included demographic and medical data, smoking, full mouth periapical radiographs, periodontal chart, occlusal analysis, and periodontal diagnosis.
Before treatment of the maxilla, the natural mandibular teeth underwent complete periodontal treatment, including preliminary cause-related therapy and additional periodontal surgery if needed.
The maxillary teeth to be extracted were gross-scaled, and the patients were instructed about proper plaque control. A recent CBCT was available before surgery in all cases.
Patients were asked to begin 0.2% CHX mouth wash several days before surgery; one hour before surgery, the patients were given premedication with 875 mg amoxicillin-clavulanate potassium (Augmentin, Smith Kline, Brentford, UK) or 600 mg clindamycin HCl (Dalacin-C, Pfizer NV/SA, Puurs, Belgium) and 8 mg dexamethasone (Aspen, Dublin, Ireland). Before the procedure, a one-minute rinsing with a 0.2% chlorhexidine solution (Tarodent mouthwash, Taro Pharmaceutical Industries, Haifa, Israel) was completed. Patients continued taking antibiotics for one week after surgery (Augmentin 875 mg × 2 per day or Dalacin 150 mg × 4 per day). The 4 mg dexamethasone treatment was carried out over two successive days. Starting the day after treatment, the patients were instructed to rinse with 0.2% chlorhexidine twice a day for two weeks. A panoramic examination was performed (Figures 1-3) at 3-time points: before treatment, immediately after implant placement, and after bridge installation. mouth periapical radiographs, periodontal chart, occlusal analysis, and periodontal diagnosis.
Before treatment of the maxilla, the natural mandibular teeth underwent complete periodontal treatment, including preliminary cause-related therapy and additional periodontal surgery if needed.
The maxillary teeth to be extracted were gross-scaled, and the patients were instructed about proper plaque control. A recent CBCT was available before surgery in all cases. Patients were asked to begin 0.2% CHX mouth wash several days before surgery; one hour before surgery, the patients were given premedication with 875 mg amoxicillinclavulanate potassium (Augmentin, Smith Kline, Brentford, UK) or 600 mg clindamycin HCl (Dalacin-C, Pfizer NV/SA, Puurs, Belgium) and 8 mg dexamethasone (Aspen, Dublin, Ireland). Before the procedure, a one-minute rinsing with a 0.2% chlorhexidine solution (Tarodent mouthwash, Taro Pharmaceutical Industries, Haifa, Israel) was completed. Patients continued taking antibiotics for one week after surgery (Augmentin 875 mg × 2 per day or Dalacin 150 mg × 4 per day). The 4 mg dexamethasone treatment was carried out over two successive days. Starting the day after treatment, the patients were instructed to rinse with 0.2% chlorhexidine twice a day for two weeks. A panoramic examination was performed (Figures 1-3) at 3-time points: before treatment, immediately after implant placement, and after bridge installation.     One surgeon (RK) completed all surgeries. After local anesthesia, intrasulcular incisions and elevation of a full-thickness flap were carried out. Subsequently, extractions were performed, and the granulation tissue was removed. All implants were placed in socket sites without using a surgical guide. Implant osteotomy was prepared by engaging the palatal and apical walls of the socket; no attempt was made to level the bone. Depending on the bone quality, the final drill was at least 1 mm smaller in diameter than the implant width to attain the desired primary stability. Implants, Lance or Seven (MIS) with a sandblasted and acid-etched surface (MIS Implants Technologies, Bar-Lev industrial zone, Israel), were inserted with a torque-controlled ratchet; primary stability was considered sufficient when the insertion torque was ≥ 30 Ncm. Implants were placed 2-3 mm palatal to the buccal wall of the extraction socket and at the level of the palatal bone or slightly apically ( Figure 4). One surgeon (RK) completed all surgeries. After local anesthesia, intrasulcular incisions and elevation of a full-thickness flap were carried out. Subsequently, extractions were performed, and the granulation tissue was removed. All implants were placed in socket sites without using a surgical guide. Implant osteotomy was prepared by engaging the palatal and apical walls of the socket; no attempt was made to level the bone. Depending on the bone quality, the final drill was at least 1 mm smaller in diameter than the implant width to attain the desired primary stability. Implants, Lance or Seven (MIS) with a sandblasted and acid-etched surface (MIS Implants Technologies, Bar-Lev industrial zone, Israel), were inserted with a torque-controlled ratchet; primary stability was considered sufficient when the insertion torque was ≥ 30 Ncm. Implants were placed 2-3 mm palatal to the buccal wall of the extraction socket and at the level of the palatal bone or slightly apically ( Figure 4).  Implant transfer adaptation was made immediately after all implants were placed ( Figure 5); control radiographs were conducted to confirm full seating. Afterward, the sockets and the ridge were augmented with a particulate freeze-dried bone allograft (FDBA) (0.25-1 mm, Life-Net, Virginia, FL, USA), which filled the residual gaps around the implants and was added in excess above the buccal bone in sites with less than 2 mm Implant transfer adaptation was made immediately after all implants were placed ( Figure 5); control radiographs were conducted to confirm full seating. Afterward, the sockets and the ridge were augmented with a particulate freeze-dried bone allograft (FDBA) (0.25-1 mm, Life-Net, Virginia, FL, USA), which filled the residual gaps around the implants and was added in excess above the buccal bone in sites with less than 2 mm of thickness. A collagen membrane (Bio-Gide Geistlisch Pharma AG, Wolhusen, Switzerland) covered the augmented areas ( Figure 6). Tension-free suturing was realized after periosteal releasing incisions of the buccal flap. The flaps were adapted using the horizontal mattress and a simple interrupted suture technique (4/0 Vicryl rapid, Ethicon, Johnson & Johnson, New Brunswick, NJ, USA) ( Figure 7). After suturing, plastic impression cap lock devices were inserted, and an impression was taken using the putty-wash one-step technique (Express, 3M ESPE Dental Products, St. Paul, MN, USA) with a closed metal stock impression tray. Finally, interarch relations were recorded.      Most sutures were gone on the 14th day follow-up visit, and the remaining ones were easily removed.

Prosthetic Protocol
One to three days after surgery, abutments were torqued at 15-20 Ncm according to implant diameter. Full seating was verified radiographically, and a prefabricated metalreinforced temporary acrylic bridge was then delivered.
The finish line of the temporary crowns was no more than 1 mm subgingival to facilitate the removal of the residual cement and facilitate plaque control. The temporary and final bridge design complied with the prosthetic principles aimed at reducing the overload, limiting the cantilever length (≤15 mm), preparing flat cusp inclines, and organizing multiple contact points centered on the crown of various teeth to increase the distribution of force, and avoiding interferences [23]. Depending on the emerging positions of the most distal implants, the provisional prostheses comprised 10-12 units; occlusal adjustments were made according to the following four principles: (1) maximum contacts in centric relation, (2) multipoint contacts during lateral and protrusive movements, (3) canine guidance; and (4) a distal cantilever without contact in any position (Figures 8 and 9). Most sutures were gone on the 14th day follow-up visit, and the remaining ones were easily removed.

Prosthetic Protocol
One to three days after surgery, abutments were torqued at 15-20 Ncm according to implant diameter. Full seating was verified radiographically, and a prefabricated metalreinforced temporary acrylic bridge was then delivered.
The finish line of the temporary crowns was no more than 1 mm subgingival to facilitate the removal of the residual cement and facilitate plaque control. The temporary and final bridge design complied with the prosthetic principles aimed at reducing the overload, limiting the cantilever length (≤15 mm), preparing flat cusp inclines, and organizing multiple contact points centered on the crown of various teeth to increase the distribution of force, and avoiding interferences [23]. Depending on the emerging positions of the most distal implants, the provisional prostheses comprised 10-12 units; occlusal adjustments were made according to the following four principles: (1) maximum contacts in centric relation, (2) multipoint contacts during lateral and protrusive movements, (3) canine guidance; and (4) a distal cantilever without contact in any position (Figures 8 and 9).     The provisional restoration was removed after six months of healing; the same impression protocol was used for the final restoration as for the provisional one. This restorative phase included using a master model with a silicon image of the marginal gingiva and new abutments for which the crown margins were located <2 mm subgingivally to allow an esthetic emergence profile and enable the removal of cement excess. The abutments were connected during the next appointment, and porcelain fused to a metal bridge with embrasure spaces that allowed the use of interproximal brushes was tried. Then, a ratchet was used to tighten the abutments to 15-35 Ncm (depending on implant diameter). Cementation of the permanent porcelain fused to a semiprecious metal bridge ( Figure 10 The provisional restoration was removed after six months of healing; the same impression protocol was used for the final restoration as for the provisional one. This restorative phase included using a master model with a silicon image of the marginal gingiva and new abutments for which the crown margins were located <2 mm subgingivally to allow an esthetic emergence profile and enable the removal of cement excess. The abutments were connected during the next appointment, and porcelain fused to a metal bridge with embrasure spaces that allowed the use of interproximal brushes was tried. Then, a ratchet was used to tighten the abutments to 15-35 Ncm (depending on implant diameter). Cementation of the permanent porcelain fused to a semiprecious metal bridge ( Figure 10) was achieved with temporary cement (Temp-Bond, Kerr Corporation, Ann Arbor, Michigan USA).

Clinical Follow-Up Examination
Supervised dental hygienists performed periodical 3-6 months supporting periodontal treatment (SPT) appointments. They included recording the plaque index, probing depth, bleeding, personal oral hygiene instructions, and scaling and root planning, if needed. The senior author (RK) performed the full periodontal evaluation once a year. Since the prosthetic design enabled full access to the implant circumference, maintenance treatment was performed without removing the prosthesis.

Outcome Measurements
Outcome measurements included: Implant success rates according to Albrektsson  Esthetic complications evaluated by the patient in terms of lip support and appearance of the artificial teeth; 3.
Functional complications, cheek and lip biting, and phonetic complaints.

Periodontal Parameters at the Last Recall Visit
• Plaque index (PI)-Using a disclosing solution, the plaque was measured at four sites per implant, and the percentage of visible plaque was calculated [22].

•
Bleeding index (BI)-A yes/no reading of bleeding within 10 s after probing at four sites for each implant was done. The bleeding index was calculated per patient (total number of bleeding sites divided by the number of implants ×4). • PD-The probing depth was measured to the nearest mm at four sites at the follow-up examinations. The mean implant probing was calculated and used for statistical analysis. • Keratinized mucosal width (KMW) was measured with a probe to the nearest mm; PD, BI, and PI were measured at the mesial, distal, buccal, and palatal sides.

Radiographic Measurements
Postoperative periapical radiographic examinations were performed at bridge installation and the annual follow-up examinations. Standardized radiographs, with the film, kept parallel (Kodak Ektaspeed plus, Eastman Kodak Co., Rochester, NY, USA) and the X-ray beam perpendicular to the implant were taken using plastic film holders (Dentsply-Rinn Corporation, Elgin, IL, USA). The bone level associated with the implants was evaluated using computerized digital radiography (Schick Technologies, New York, NY, USA) ( Figure 11). Radiographic evaluations were performed by an independent examiner (AR).
Evaluations were performed by measuring the distance between the marginal bone apical to the implant shoulder and the implant shoulder, which served as a reference level (RL). The distance from the RL to the first bone-to-implant contact was measured on the mesial and distal sides of the implant. Radiographic distortion was calculated by dividing the radiographic implant length by the actual implant length. Bone loss was measured at 1, 3, and 5 years after prosthetic delivery. The grafting material induced a heavy masking effect on the peri-implant bone levels on the radiographs taken immediately after surgery. Subsequently, it was necessary to use a later time reference to compare the MBL of the axial and tilted implant, and the 1-year recall after bridge installation was decided to serve as the initial reference milestone. The difference (ΔH) between the final (5 years) and the initial measurements (1 year) was calculated. Since marginal bone loss is time-dependent, both the annual bone loss and the raw (total) MBL at the last follow-up were determined.
When the marginal bone was coronal to the implant shoulder, it was considered a zero-bone loss; only changes apical to the implant shoulder were measured. Implant angulation was measured by tracing two lines, one through the occlusal plane and the other parallel to the long axis of the implant (Figure 11). Angulation between each implant and the occlusal plane was calculated by subtracting the intersection angle from 90 degrees. The average angulation of the axial and tilted implants and the average annual mesial and distal bone loss for the axial (A) and tilted (T) implants were calculated. Figure 11. Measurement of the marginal bone level and the angle of the tilted implant with dedicated software. After calibration, the implant shoulder was used as a reference level (RL), and the distance from the RL to the first bone-to-implant contact was measured. The angle of the tilted implants was measured by tracing lines through the occlusal plane and parallel to the long axis of the implants. Radiographic distortion was calculated by dividing the radiographic implant length by the actual implant length. Bone loss was measured at 1, 3, and 5 years after prosthetic delivery. The grafting material induced a heavy masking effect on the peri-implant bone levels on the radiographs taken immediately after surgery. Subsequently, it was necessary to use a later time reference to compare the MBL of the axial and tilted implant, and the 1-year recall after bridge installation was decided to serve as the initial reference milestone. The difference (∆H) between the final (5 years) and the initial measurements (1 year) was calculated. Since marginal bone loss is time-dependent, both the annual bone loss and the raw (total) MBL at the last follow-up were determined.
When the marginal bone was coronal to the implant shoulder, it was considered a zero-bone loss; only changes apical to the implant shoulder were measured. Implant angulation was measured by tracing two lines, one through the occlusal plane and the other parallel to the long axis of the implant (Figure 11). Angulation between each implant and the occlusal plane was calculated by subtracting the intersection angle from 90 degrees.
The average angulation of the axial and tilted implants and the average annual mesial and distal bone loss for the axial (A) and tilted (T) implants were calculated.

Statistical Analysis
The primary outcome variables were the survival and success rates of the axial and tilted implants and the MBL between the 1-and 5-year recall. The ANOVA test with repeated measures was used in a mixed statistical analysis model, taking into account that each patient had several implants. The model compared the MBL rate between implants displaying an angulation of <15 • and ≥15 • , implants supporting or not supporting a cantilever, sites with and without KTW, smokers and nonsmokers, and implant location. The SPSS version 25.0 (IBM, Armonk, NY, USA) was used for the statistical analysis.

Demographic Data
Twenty-three consecutive patients, eleven men and twelve women, met the inclusion criteria. The mean age was 65.34 ± 10.46 years at implant placement (range 46-90). Eight patients were smokers (two of them ≥ 10 cigarettes/day) ( Table 1). Eighty-four straight implants and forty-six tilted implants were placed in the postextraction sockets of the maxilla. The first molar was the most common site for the tilted implants (52%). The distribution of implants according to type, diameter, and length is shown in Table 2. Of the 130 implants (MIS Dental Implant Technologies, Bar-Lev Industrial Park, Israel), 112 were Lance (86.2%) and 18 were Seven (13.8%). The most used implant length was 16 mm (n = 93, 71.5%) for both the axial and tilted implants; the most frequent implant diameter was Ø 3.75 mm (67/130, 51.5%). The schematic outline of the study is described in Table 3.

Implant Survival and Success Rate
No implant was lost during the first and third years, and the survival rate was 100%. One tilted implant failed afterward, and the 5-year survival rate of the straight and tilted implants was 100% and 97.8% (p > 0.05), respectively; our hypothesis was confirmed. The prosthetic survival rate was 100%. The total (straight + tilted) 5-year success rate, according to Albrektsson et al. [24], was 94.6%.

Complications Biological Complications
Subnasal and suborbital hematomas occurred in 10 patients (43.4%) during the first week postsurgery. The mechanical, functional, and biological complications data are summarized in Table 4.
Peri-implantitis involving seven implants (two of them tilted) was diagnosed in three patients, of whom one was a nonsmoking 64-year-old woman that had four involved implants (one tilted implant failed). Those implants presented redness, edema, mucosal enlargement, deep periodontal pockets with bleeding upon probing, and progressive bone loss. The implants were treated with nonsurgical therapy and maintained function ( Table 4). The failed implant was extracted, and the abutment hole was closed using composite resin. The bridge was recemented on five implants.
In the case of decementation, the temporary cement was replaced with Cem-Implant (B.J. M laboratories, Ann Arbor, MI, USA). Two female patients complained about the height of the crowns of their prosthesis; over time, they became accustomed to the change, and no intervention was needed (Table 4). Two other female patients experienced lip or cheek biting; this was solved by minimal selective grinding of the lower relevant teeth to increase the overjet dimension (Table 4).

Marginal Bone Loss
The 5-year MBL of the mesial and distal sides of the axial implants was 0.42 ± 0.67 mm and 0.59 ±1.01, respectively; for the tilted implants, it was 0.37 ± 0.68 mm and 0.34 ±0.62, respectively. The differences between the straight and tilted implants were not statistically significant; p was 0.67 and 0.09 for the mesial and distal sides, respectively, and our hypothesis was confirmed. For the straight implants (n = 84), the mean annual MBL rate was 0.08 ± 0.13 mm for the mesial side and 0.12 ± 0.20 mm for the distal one (Table 5); for the tilted implants (n = 45), it was 0.07 ± 0.14 mm and 0.07 ± 0.12 mm for the mesial and distal sides, respectively (Table 5). One tilted implant did not survive the 5-year follow-up examination due to severe peri-implantitis. The difference in the MBL annual rates between the axial and tilted implants was not statistically different (p = 0.629 for the mesial side and p = 0.083 for the distal one).  A comparison of the mean MBL rate (mm) as a function of different covariates is presented in Table 5; implant position, implants with or without a cantilever, and smoking status did not affect the MBL rate.

Discussion
In the present study, the 5-year total survival rate of immediate implants supporting a maxillary prosthesis relying on four or six immediately loaded axial and tilted implants in periodontally compromised patients was 99.2%. This finding is like or slightly superior to the previous reports on this type of restoration [27][28][29].
Most studies dealt with screw-retained restorations [30]; in contrast, the restorations of this study were cemented. The advantages of a cemented restoration include the compensation of improperly inclined implants, easier achievement of passive fit, and an intact occlusal table without a screw-access hole resulting in better control of the occlusion [30]. On the other hand, cement excess is related to more biological complications [31]. This was the reason for placing the crown margins <2 mm subgingivally, enabling control and removal of the cement excess. Previous data showed that undetected cement excess increased 3-4-fold when the restoration margin was 2-3 mm subgingivally compared to 1 mm [31].
Our findings are in line with previous studies showing no correlation between the implant location/angulation and MBL or survival rates [27,32,33].
Most authors reported an MBL ranging from 1.30 to 1.72 mm at the 3-to 5-year follow-up for the axial and tilted implants supporting a screw-retained fixed complete denture prosthesis (FCDP) [11,12,28,29]. The 5-year MBL of the present implants placed in patients diagnosed with advanced periodontitis, extensive bone loss, and terminal maxillary dentition was 0.41 mm and 0.50 on the mesial and distal sides, respectively, with no difference between the axial and tilted implants. The current periodontal and hygiene-oriented bridge construct may have contributed to the superior bone results; it is probably due to the morphology of the final restorations that enabled reasonable plaque control and full access to the whole circumference of the abutment/crown interface with electrical/manual brushes. Noteworthy, all patients of this cohort adhered to an individualized SPT, which correlated with a lower incidence of implant and bone loss [34].
Peri-implantitis involved only 13% of the patients and 5.4% of the implants compared to 22% of the patients in another study that used a computer-guided surgical protocol (Nobel Guide) and immediately loaded implants with a screwed-retained prosthesis followed for 5 years [11].
Our technical complications rate was lower than that presented by Lopes et al., who reported mechanical complications in 81.9% of patients, of which fractured prosthesis and abutment loosening contributed to 59.4% and 60.3%, respectively [11]. In line with the superior results of the current study using cemented restorations, a systematic review by Weber et al. [35] in partially edentulous patients concluded that the success rates of screw-retained and cement-retained implant-supported FDPs after a follow-up period of 6 years were 93.2% for the cemented and 83.4% for the screw-retained restorations.
The low PD (4.20 ± 1.25 mm) measured during the last follow-up examination indicated the presence of a healthy peri-implant soft tissue; this feature is most likely related to the high level of maintenance and the characteristics of the prostheses that facilitated the maintenance of the implants, as already reported in our previous studies [9,36].
Alveolar ridge volume has been found to improve with a range of different bone substitutes [37]; preclinical data [38,39] have demonstrated that the use of allografts covered with a collagen membrane may limit vertical bone loss. Limited resorbability of the grafting material can also be advantageous, as it minimizes resorption of the buccal bone [40]. In the present study, we used an FDBA inside and outside the sockets covered by a resorbable collagen membrane. This fact may have contributed to the minimal MBL and high success rate of these implants.
The presence of a distal cantilever did not increase the MBL rate of the tilted implants. This is in line with a recent meta-analysis by Freitas et al. that did not find a correlation between the presence of a cantilever and MBL in fixed partial [41] or full-arch implantsupported restorations [42].

1.
Only patients who had been followed for 5 years were included, and the number of patients was therefore limited; 2.
Longer follow-ups are warranted, since biological complications are time-dependent; 3.
The present study was retrospective and covered all consecutive patients treated with the all-on-four-six protocols by a single periodontist; thus, no power was calculated. Nevertheless, a recent systematic review included all studies with a minimum of 20 patients [15].
The strengths of the study were:
The treatment protocol was performed according to the state-of-the-art knowledge, as mandibular teeth were without active periodontal disease; 3.
All implants were placed in the maxilla, using the same implant system, bone substitutes, and resorbable membrane in highly compliant patients.

Conclusions
Within the limitations of this study, we can conclude that our results indicate that if the prerequisites for immediate loading such as high primary stability of ≥30 Ncm and splinting of the implants via a provisional prosthesis and the use of bone level implants with a sandblasted and acid-etched surface are fulfilled: 1. Full-arch fixed restorations supported by a combination of axial and tilted implants can be a viable treatment option to rehabilitate the terminal dentition of the severe generalized periodontitis patients; 2. Neither the implant survival rate nor the peri-implant marginal bone loss seems to be affected by the inclination of the implants with respect to the occlusal plane; 3. The treatment predictability and clinical outcomes of axial and tilted implants seem comparable.
Further prospective controlled studies using the present prosthetic protocol with more extended follow-up periods are warranted to evaluate this type of treatment in patients with periodontitis.  Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki; it was approved by the Ethics Committee of Tel Aviv University, approval number 0000030-1.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data supporting reported results can be found in the tables.

Acknowledgments:
The authors would like to thank Serge Szmukler-Moncler for reviewing and editing the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.

MBL
Marginal bone loss TI Tilted implant KMW Keratinized mucosal width CBCT Cone-beam computed tomography FDBA Freeze-dried bone allograft SPT Supporting periodontal treatment RL Reference level FCDP Fixed complete denture prosthesis