Complication, vertical bone gain, volumetric changes after vertical ridge augmentation using customized reinforced PTFE mesh or Ti‐mesh. A non‐inferiority randomized clinical trial

Abstract

Objective

The aim of this non‐inferiority randomized clinical trial was to compare the surgical and healing complications, vertical bone gain, and volumetric bone changes after vertical ridge augmentation using two different approaches: customized Ti‐reinforced d‐PTFE mesh versus customized CAD/CAM Ti‐mesh.

Materials and Methods

Fifty patients with vertical bone defects were randomly treated with Ti‐reinforced d‐PTFE mesh (control group) or CAD/CAM Ti‐mesh (test group) and a mix of autogenous bone and deproteinized bovine bone matrix. Surgical and healing complication rates (SCR‐HCR), vertical bone gain (VBG), regenerated bone volume (RBV), and regeneration rates (RR and ERR) were recorded and analysed [significance level (α) of 0.05].

Results

Of the 50 patients, 48 underwent bone augmentation surgery. SCR were 4% and 12% in PTFE and Ti‐mesh, whereas HCR were 12.5% and 8.3%. VBG were 5.79 ± 1.71 mm (range: 3.2–8.8 mm) in the PTFE group and 5.18 ± 1.61 mm (range: 3.1–8.0 mm) in the Ti‐mesh group (p = .233), whereas RBV were 1.46 ± 0.48 cc and 1.26 ± 0.55. RR was 99.5% and 87.0%, demonstrating a statistically significant difference (p = .013). Finally, the values related to pseudo‐periosteum, bone density, and implant stability were similar in the two study groups. Osseointegration rates were 98.2% and 98.3%.

Conclusions

This study confirmed the non‐inferiority of customized CAD/CAM titanium meshes with respect to reinforced PTFE meshes in terms of surgical and healing complications. Although PTFE meshes showed higher vertical bone gain and regeneration rates than Ti‐meshes, no significant differences were found.

1. INTRODUCTION

Vertical ridge augmentation (VRA) procedures are essential in oral and maxillofacial surgery to restore lost bone volume resulting from trauma, infection, or other pathologies. In recent years, some prominent and novel devices based on the digital 3D approach have emerged for guiding and supporting bone regeneration: customized CAD/CAM titanium meshes (Chiapasco et al., 2021) and titanium‐reinforced polytetrafluoroethylene (d‐PTFE) perforated membranes (Cucchi et al., 2022).

The first barrier device was developed by means of computer‐aided design/computer‐aided manufacturing technology based on 3D titanium laser sintering. This made it possible to create a highly accurate three‐dimensional framework to precisely fit the contours and the morphology of all kinds of bone defects, achieving high implant survival rates and low peri‐implant bone loss (Hartmann et al., 2022).

The second barrier device was developed as a combination of the traditional features of titanium mesh and PTFE membranes and it is known as Reinforced PTFE Mesh (RPM). It incorporates circular macropores within its PTFE composition, enabling direct communication between the surgically created space and the periosteum above. This feature enhances revascularization and facilitates the influx of progenitor cells. The combined effects of improved vascularization, progenitor cell recruitment, and extracellular matrix formation facilitated by macropores can lead to faster bone maturation. This means that the regenerated bone becomes denser and stronger more quickly, achieving complete regeneration in all sites with baseline deficiencies of <5 mm, in 95.6% of sites with 5–8 mm deficiencies, and in 89.4% of sites with >8 mm deficiencies (Urban et al., 2021).

Customized titanium meshes have gained popularity due to their adaptability to the patients' specific anatomical. These meshes are durable, rigid, and provide excellent structural support for bone grafts, reducing the risk of graft displacement. Non‐resorbable PTFE membranes are characterized by their biocompatibility and barrier function, which aid in protecting the graft material from soft tissue invasion. They are flexible and relatively easy to handle during surgical procedures.

Despite the advantages of both techniques, complications can arise impacting patient outcomes and satisfaction. Both titanium meshes and PTFE membranes may serve as potential sites for infection, that can lead to prolonged healing times, to partial or complete graft failure, peri‐implant bone loss, and compromised implant survival rates (Poli et al., 2022; Urban et al., 2023).

Many authors (Chiapasco et al., 2021; Cucchi et al., 2022; Hartmann & Seiler, 2020; Lizio et al., 2022; Seiler et al., 2018) have published data about complication rates, vertical bone gain, and volumetric changes after alveolar ridge augmentation using customized CAD/CAM mesh or reinforced PTFE mesh, but these studies had methodological flaws, biases, or limitations that reduced the strength and the quality of evidence (Urban, Montero et al., 2019; Urban, Nagy et al., 2019).

The aim of this randomized clinical trial was to investigate the non‐inferiority of customized CAD/CAM titanium mesh versus titanium‐reinforced d‐PTFE mesh, in regard to the complication rates, the vertical bone gain, and the volumetric bone changes after vertical ridge augmentation.

2. MATERIALS AND METHODS

2.1. Study design

The current investigation was structured as a non‐inferiority, double‐blind, parallel‐group randomized clinical trial. The study conforms to the guidelines outlined in the CONSORT statement (http://www.consort‐statement.org).

This study was conducted in accordance with the ethical principles set forth in the Declaration of Helsinki (2014). Ethical approval for the study protocol (protocol CMF 01/2019) was obtained from the Central Emilia Wide Area Ethical Committee of the Emilia‐Romagna Region (CE‐AVEC, study code CE 19143), and the protocol was registered on ClinicalTrials.gov (NCT04257097).

Ensuring that all patients were well‐informed about the surgical protocol, with a comprehensive explanation of both its advantages and disadvantages, and an explicit assessment of the benefit–risk ratio, were pivotal procedural prerequisites. All patients received a written consent form along with a detailed document outlining the treatment they were scheduled to undergo.

The inclusion criteria for the study encompassed individuals who were over the age of 18, experienced partial edentulism (≤5 teeth) in either the maxilla or mandible, had a vertical bone defect (VBD) >3 mm, requiring three‐dimensional bone augmentation for prosthetically guided implant placement, and had provided informed consent for experimental treatment and for processing their personal data.

Conversely, individuals who met any of the following exclusion criteria were not considered eligible for the study: poor oral hygiene, untreated periodontal disease, a smoking habit of >10 cigarettes per day, substance abuse, pregnancy, the presence of odontostomatological and/or systemic infections, underlying metabolic, liver, kidney, or autoimmune diseases, a history of head and neck radiotherapy within the past 5 years, and individuals currently undergoing immunosuppressive therapy and/or those who were immunocompromised.

The 50 enrolled patients were randomly assigned to two study groups using a computer‐generated randomization sequence. Twenty‐five patients were assigned to the control group, also known as the ‘PTFE group’, and were treated with a Ti‐reinforced d‐PTFE mesh (RPM, Osteogenics) in conjunction with a pericardium collagen membrane (Vitala, Osteogenics). The remaining 25 patients constituted the test group, referred to as the ‘Ti‐mesh group’, and were treated with a custom‐made titanium mesh (Yxoss CBR, Reoss) covered with a native collagen membrane (Biogide, Geistlich) (Figure 1). All patients received a combination 50:50 autogenous bone chips and deproteinized bovine bone matrix (DBBM) (Bio‐oss, Geistlich) as grafting material. Treatment allocations were not disclosed to the patients, clinicians, or the statistician (blinded design).

FIGURE 1.

Guided bone regeneration procedure performed on the 50 enrolled patients of the two study groups (25 patients treated using titanium‐reinforced PTFE membranes, 25 patients treated using customized titanium meshes). Patients with complications were represented in black and white: one flap perforation, two exposures, one abscess in PTFE group; one flap perforation, one exposure, one abscess in Ti‐mesh group. The grey pictures referred to patients with complications.

Sample size determination was based on the primary outcome specified in the study protocol, which aimed to assess healing complications associated with vertical ridge augmentation. Power analysis was conducted to establish a non‐inferiority margin of 10%, a statistical power of 80%, and a significance level of 5%. To provide adequate support for the hypothesis of non‐inferiority in the ti‐mesh group (test) compared with the PTFE group (control), a minimum of 17 patients per group was required. Accounting for potential dropouts, the sample size for this study protocol was increased to 25 patients per group, totalling 50 patients. The research hypothesis was that the customized CAD/CAM meshes (test group) were not inferior to reinforced d‐PTFE mesh (control group) in terms of healing complication rates, so that the complication rate in the Ti‐mesh group was equal to or lower than the complication rate in the PTFE group +10%.

2.2. Digital workflow

All patients underwent cone beam computed tomography (CBCT) scans to generate the necessary Digital Imaging and Communications in Medicine (DICOM) files for the digital planning of bone augmentation at T0.

The virtual bone augmentation process for all defects was overseen by a blinded clinician (G.C.) well‐experienced in digital planning and bone augmentation.

In the PTFE group, the DICOM files were uploaded on a platform designated by the manufacturer (Biotec srl, Dueville) and used to generate the 3D reconstruction of the bone defect. After the virtual placement of the implants in the correct 3D positions, the bone volume was virtually reconstructed, and a virtual customized mesh (named ‘replica’) was designed atop this virtual reconstruction. The STL file was then overlaid on the DICOM files within BTK‐3D and shared with the clinician. After the final approval of the project, both the 3D bone model and the replica of the mesh were produced using translucent acrylic resin (Vitra 430 resin, DWS System srl, Thiene).

In the Ti‐mesh group, the DICOM files of pre‐operative CBCT and the STL files of bone augmentation were sent to the CAD/CAM manufacturer (MyReoss, Reoss) for the creation of a designed 3D virtual model replicating the bone defect and the customized mesh. Any necessary modifications were made by the clinician using specialized software and email contacts. After receiving final approval from the clinician, a precise and accurate customized mesh was manufactured for each patient using CAD/CAM technology through laser sintering (Figure 2).

FIGURE 2.

(a–h): Representation of virtual bone augmentation process in two different cases. (a, e) overall view of the bone augmentation process, carried out on DICOM files extracted from pre‐operative CBCT. (b, f) Initial bone defect. (c, g) Virtual bone augmentation. (d, h) Virtually planned customized mesh.

Based on the randomization sequence, that was known only by the second surgeon (S.B.), either the 3D bone model and the mesh replica (Cucchi et al., 2022) or the customized CAD‐CAM titanium mesh (Seiler et al., 2018) were ordered and manufactured before the surgery; consequently, a sealed envelope for each subject allocation was prepared and kept closed until the patients had to receive the medical device for VRA.

All surgical procedures were performed by the same experienced surgeon (A.C.), who remained unaware of the patient's allocation group until the non‐blinded clinician (S.B.) disclosed this information during the surgery, just before the fixation of the device.

2.3. Surgical protocol

Premedication was administered 1 h prior to the procedure (T0). This included antibiotics (2 g amoxicillin/clavulanic acid and 500 mg metronidazole), anti‐inflammatories (20 mg piroxicam), and oral sedatives (2 mL delorazepam and 2 mL diazepam). To reduce as much as possible the number of bacteria in the mouth, to minimize the risk of contamination, and consequently to decrease the risk of infection, patients underwent a thorough oral rinsing protocol with three different solutions (Cucchi et al., 2022): povidone‐iodine for 1 min, hydrogen peroxide for 2 min, and a 0.2% chlorhexidine mouthwash for 3 min. Sterile drapes were applied around the surgical site, and the tongue and extraoral skin were cleansed with povidone‐iodine. Local anaesthesia was administered via local infiltration using 4% articaine with 1:100,000 adrenaline.

The second surgeon (S.B.) prepared the barrier device according to the allocation group of each patient prior to surgery. If the patient belonged to PTFE group, both the STL model and the mesh replica were disinfected for 30 min using a 70% alcohol solution. The reinforced d‐PTFE mesh was manually shaped according to the STL model of the augmented bone and the mesh replica. This shaping process involved aligning the RPM with the mesiodistal extent of the defect and positioning it between the 3D bone model and the replica. A 15C blade was used to shape the RPM in accordance with the replica's borders and to match it to the augmented bone model. Subsequently, the device was placed in a previously sterilized package, which was immediately sealed to maintain its sterility. The customized mesh for patients in the Ti‐mesh group was delivered non‐sterile in double‐peel bags which had to be sterilized before use. The double sealed bags containing the mesh were removed from the carton and placed directly in the autoclave and sterilized at 135°C for 20 min in confined water vapour.

All surgical procedures of vertical ridge augmentation (T0) and the following computer‐guided implant surgery (T1) were performed by the same first surgeon (A.C.) with expertise in both vertical ridge augmentation methods.

The surgical procedure began with a horizontal incision (mid‐crestal in the mandibular and para‐crestal in the maxilla), followed by two vertical mesial and distal releasing incisions on the buccal side and an oblique mesial releasing incision on the lingual side. Particular care was taken to prevent flap laceration and perforation. A full‐thickness flap was delicately raised while preserving vital structures such as the mental nerve, infraorbital nerve, nasopalatine nerves, major palatine artery, nasal cavities, or maxillary sinus, depending on the treated region.

On the buccal side, flap mobilization involved a longitudinal periosteal incision in the vestibular fornix and a superficial partial‐thickness dissection of the mucous component from the muscle‐periosteal component. In the maxilla, in case of severe vertical defects >4 mm or difficulty for flap mobilization or scarred periosteum due to previous failed surgeries, the Bichat's fat pad was also used as a pedicled or free adipose graft to reduce the risk of exposure or infection (Cucchi et al., 2023). On the lingual side, a full‐thickness mucoperiosteal flap was raised up to the mylohyoid line, and subsequently released by detaching the mylohyoid muscle insertion from the inner part of the flap. Autogenous bone chips were harvested using the bone scraper technique (SafeScraper Twist, Meta), followed by cortical bone perforation to promote angiogenesis. A 3‐mL sample of peripheral venous blood was collected from the median cubital vein using a sterile syringe. A 50:50 mixture of autogenous bone and xenograft was created using deproteinized bovine bone matrix (DBBM) (Bio‐Oss, Geistlich), with peripheral venous blood added as previously described.

In PTFE group, the pre‐shaped and modelled d‐PTFE mesh, supported by the mesh replica, was filled with the grafting material and positioned in the surgical site. The mesh replica served as a scaffold to maintain the correct shape of the d‐PTFE mesh and was secured in place using osteosynthesis screws (Pro‐Fix Fixation System, Osteogenics) and/or titanium tacks (McBio s.r.l.). The presence of holes on the replica's surface facilitated RPM stabilization in the digitally planned position with screws and/or tacks before removing the replica from the surgical site, thus reducing the risk of fixation errors. Subsequently, the RPM was covered with a resorbable collagen membrane (Vitala, Osteogenics), which had been hydrated with sterile saline solution 30 min prior to fixation to align the start of the hydrolytic degradation process with the anticipated resorption time of the pericardium membranes and that of the native collagen membranes. This membrane was shaped and fixed over the RPM with tacks to enhance its stability (Figure 3).

FIGURE 3.

Augmentation procedure performed using titanium‐reinforced PTFE mesh in maxilla. (a, b) Pre‐operative intraoral clinical view. (c) Augmented 3D bone model and replica of the mesh. (d) PTFE mesh modelled and tailored, placed between the 3D bone model and the replica. (e) Pre‐modelled PTFE mesh, supported by the replica, filled with the graft mixture. (f) Filled PTFE mesh placed in the surgical site, using the replica as scaffold. (g) Filling of the PTFE mesh using additional graft, before placing the vestibular screws. (h) PTFE mesh fixed in the surgical site. (i–k) Pedicled BFP released and advanced, without tension, to cover the whole area. (j–l) Tension free primary closure obtained with a combination of horizontal mattress sutures and multiple interrupted sutures.

In the Ti‐mesh group, the customized titanium mesh (Yxoss CBR, Reoss) was similarly used. The grafting material was placed into the device, then positioned within the surgical site, and fixed with 3–4 osteosynthesis screws. Following this, a resorbable collagen membrane (Biogide, Geistlich) was applied to cover the mesh and fixed using titanium tacks (Figure 4).

FIGURE 4.

Augmentation procedure performed using customized titanium mesh in maxilla. (a, b) Pre‐operative intraoral clinical view. (c, d) Virtual rendering of customized titanium mesh. (e) Customized titanium mesh and mixture of autogenous bone, xenograft and peripheral venous blood. (f) Customized titanium mesh filled with the graft mixture. (g, h) Clinical view of the bone defect. (i–k) Customized titanium mesh fixed in the surgical site. (j) Customized titanium mesh covered with a resorbable collagen membrane. (l) Pedicled BFP released and advanced, without tension, to cover the whole area.

The surgical flaps were meticulously adapted to ensure complete coverage of the augmented area with tension‐free primary closure, using a combination of horizontal mattress sutures and multiple interrupted sutures (Resorba Glycolon, Osteogenics).

After the surgery, patients followed a comprehensive postoperative medical protocol. This included a regimen of antibiotics (1 g amoxicillin/clavulanic acid and 250 mg metronidazole every 8 h for 7 days) and mouth rinsing (0.2% chlorhexidine mouthwash for 2 min, three times a day for 14 days). Postoperative instructions included ice therapy for 72 h, a soft and cold diet for 14 days, abstinence from smoking for 7 days, avoidance of rinsing and spitting for 72 h, refraining from any stress or physical exertion in the surgical site for 20 days, and abstaining from chewing in the area for 6 months. The sutures were removed 14–21 days after the surgery.

During the entire healing period, removable partial dentures were not used to prevent interference with surgical site healing. Patient evaluations were performed at 14 days (suture removal), 30 days, 3 months, and 6 months after reconstructive surgery to monitor for potential healing complications.

A second CBCT scan was done at 6 months for mandibular defects and at 9 months for maxillary defects, to allow for an adequate bone formation and mineralization. The data from this 3D CT‐scan were used to evaluate bone augmentation and to plan the computer‐guided implant surgery (Navimax, Biomax). Implants were planned 1 mm under the newly‐formed bone and in relation to the prosthetic restoration, to have a prosthetically‐guided implant placement (Figures 5 and 6).

FIGURE 5.

Planning of computer‐guided implant surgery after GBR using PTFE mesh. (a–d) CBCT performed 9 months after the bone augmentation procedure, revealing the regenerated bone volume. (e–h) Prosthetic‐guided digital implant planning. Implants were virtually placed 1 mm subcrestally with regard to the newly formed bone.

FIGURE 6.

Planning of computer‐guided implant surgery after GBR using customized Ti‐mesh. (a–d) CBCT performed 9 months after the bone augmentation procedure, revealing the regenerated bone volume. (e–h) Prosthetic‐guided digital implant planning. Implants were virtually placed 1 mm subcrestally with regard to the newly formed bone.

Subsequently, at the re‐entry surgery (T1), the barrier device (reinforced d‐PTFE membrane or customized 3D mesh) was removed along with all osteosynthesis screws and tacks. After clinical evaluation of newly‐formed bone tissue and core drilling of bone biopsy, the surgical guide/template was placed and fixed in the planned position. The implant sites were prepared using progressive twist drills and planned tapered implants were placed in the prosthetically‐driven position and axis (T3 implants, Zimvie). Sutures were applied to close the treated site and to have a submerged healing of implants for 3–4 months according to the implant stability (Figures 7 and 8).

FIGURE 7.

Re‐entry surgery performed after guided bone regeneration using titanium reinforced PTFE mesh. (a) Pre‐operative endoral radiograph (T0). (b) Post‐operative endoral radiograph (T1). (c, d) Intra‐oral clinical view before reopening. (e, f) Reopening of the treated area: Exposure of the PTFE mesh. (g, h) Removal of the PTFE mesh: Clinical view of regenerated bone. (i–k) Implants site preparation and implants placement using surgical guide template. (j–l) Implants placed in regenerated bone.

FIGURE 8.

Re‐entry surgery performed after guided bone regeneration using customized titanium mesh. (a) Pre‐operative endoral radiograph (T0). (b) Post‐operative endoral radiograph (T1). (c, d) Intra‐oral clinical view before reopening. (e, f) Reopening of the treated area: Exposure of the Ti‐mesh. (g, h) Removal of the titanium mesh: Clinical view of regenerated bone. (i–k) Implants site preparation and implants placement using surgical guide template. (j–l) Implants placed in regenerated bone.

After 3–4 months (T2), the implants were reopened and uncovered using healing screws, confirming implant osseointegration by the application of 25 Ncm manual reverse torque and managing the peri‐implant soft tissue with connective tissue graft or strip gingival graft (Urban et al., 2015; Urban, Montero et al., 2019; Urban, Nagy et al., 2019).

2.4. Data collection

2.4.1. Population data

All clinical data were collected by a blinded examiner (L.T.), using a specific case report form (CRF). The following population data were recorded: patient identification code; study group; age and gender; medical and dental anamnesis; smoking habit; periodontitis; and diabetes. Then, the following defect data were collected: region of interest (maxilla/mandible; anterior/posterior), extent of defect (number of missing teeth), dimension of defect (small, medium, large), type of defect (vertical or combined horizontal/vertical), Kennedy's classification, type of periosteum (native/scarred, based on previous surgical treatments), and bone density (high, medium, low).

2.4.2. Surgical and healing complications

Surgical complications were assessed from T0 to T1 and included the following issues: (a) failure to achieve primary intention closure of surgical flaps, (b) flap damage such as lacerations or perforations, (c) severe bleeding or haemorrhaging, and (d) neurological damage, such as temporary or permanent paraesthesia/dysaesthesia/hypoaesthesia/anaesthesia.

Healing complications were evaluated from T0 to T1, including 4 different classes as described by Fontana et al., 2011: membrane exposure <3 mm, no purulent exudate (class I); membrane exposure >3 mm, no purulent exudate (class II); membrane exposure, with purulent exudate (class III); and abscess without membrane exposure (class IV). The healing complications were also classified according to Hartmann's classification (Hartmann et al., 2019): punctual exposure (class A), exposure of one tooth width (class B), a complete exposure (class C), and no exposure (class D).

Considering the number of sites with exposures and sites with infection or suppuration or abscess with respect to the total number of treated sites, it was possible to calculate the exposure rate (%) and the infection rate (%) for each study group. Finally, based on the consequences of healing complication on the augmented sites, the bone graft loss was classified as (i) absent, (ii) partial, or (iii) total.

2.4.3. Linear measurements

For each patient, the vertical bone defect (VBD) was calculated as a linear measurement in millimetres (mm) using dedicated software (ImageJ, NIH) on the cross‐sectional image of the pre‐operative CBCT with maximum defect with respect to the line of planned bone volume, as explained in Figure 2. The distance from the point of maximum defect to the adjacent tooth was recorded and used as the reference in the post‐operative CBCT. The corresponding cross‐sectional image was used for measuring the vertical bone gain (VBG), following a similar procedure for the measurement.

2.4.4. Volumetric measurements

During the digital planning of bone augmentation, the dedicated software (BTK Opera, Biotec srl, Vicenza, Italy) permitted an exact calculation of the planned bone volume (PBV), expressed in cubic centimetres (cc), on the STL model from pre‐operative CBCT. Similarly, using the STL model from the post‐operative CBCT, the software made it possible to measure the augmented bone volume (ABV) and the regenerated bone volume (RBV). ABV was calculated including all the space under the barrier device; RBV was obtained measuring only the mineralized bone tissue, excluding the pseudo‐periosteum (non‐mineralized tissue).

For each patient, the post‐operative CBCT were imported into the software (BTK Opera by RealGuide, 3diemme). Using cropping tools, the area where the guided bone regeneration procedure had previously been performed, was selected using a manual selection. This phase was made by a calibrated operator (L.T.) on cross‐section of post‐operative CBCT and it really increased the accuracy for distinguishing the effective mineralized tissue and non‐mineralized tissue. Subsequently, in the software, the pre‐operative STL model was imported as the ‘bone surface’ and overlaid onto the CBCT using either an assisted matching or point matching method. Similarly, the STL model of the mesh was imported and overlaid as a ‘bar’. Through the segmentation tool, the STL model of the post‐operative CBCT was obtained. This model was selected and modified in the ‘3D objects’ section. During this phase, the overlapping STL files and the post‐operative CBCT were used, and cropping tools were used to obtain STL files of the augmented bone volume (ABV) and the regenerated bone volume (RBV). The obtained STL files were visualized in the ‘3D objects’ section, where their volumes in mm3 were calculated and then approximated and converted into cubic centimetres (cc).

The difference between the ABV and the RBV was defined as lacking bone volume (LBV).

The regeneration rate (RR%) was calculated as the ratio between regenerated bone volume and planned bone volume (RR% = RBV/PBV), whereas the effective regeneration rate (ERR%) was calculated as the ratio between regenerated bone volume and augmented bone volume (ERR% = RBV/ABV) (Figure 9).

FIGURE 9.

Process of volumetric measurements. (a) STL model obtained from post‐operative CBCT. (b) Superimposition of the post‐operative on the pre‐operative STL files. (c, d) Selection, cropping, and calculation of the augmented bone volume (ABV) and the regenerated bone volume (RBV).

2.4.5. Clinical data

The pseudo‐periosteum was clinically assessed at T1, using a UNC‐15 periodontal probe and classified (Cucchi et al., 2019) as: type I (no pseudo‐periosteum or <1 mm), class 2 (between 1 and 2 mm), and class 3 (irregular pseudoperiosteum or >3 mm).

Bone density was evaluated at T1, using a calibrated probing force of 20–30 g, based on the resistance of the newly formed bone to probe penetration. Assessment took place at the top of the alveolar ridge in a vertical direction and buccally in a horizontal direction, specifically at the planned implant site. Bone density was categorized as: high density, medium density, or low density.

Implant stability was assessed using a surgical handpiece at T1, categorizing implants into three groups based on torque levels: low (<15 N/cm), medium (between 15 and 35 N/cm), and high (>35 N/cm). Finally, implant osteointegration was evaluated at T2 by applying a counter torque of 25 N/cm.

2.5. Statistical analysis

Excel data collection form and a data management system were used (Microsoft Excel 2011; Windows, ver. 14.0.0; Microsoft Corp.). All data were entered by a single blinded operator. Prior to entry, all data were evaluated in terms of accuracy and completeness. The mean, median, standard deviation (SD) and 95% confidence interval (95% CI) were reported for each continuous variable. This statistical analysis, calculated at the patient level, was based on the hypothesis (H1) that Ti‐mesh (test group) would not be inferior to reinforced PTFE mesh (control group) in the incidence of healing complications (primary outcome). A ‘Per‐Protocol analysis’ was chosen to describe the collected data and to explore the results, so only patients that followed strictly the study protocol were included in the statistical analysis.

The non‐inferiority test was performed for Healing and Surgical complication only (one‐sided 95% confidence interval approach), specifying the greatest difference that was clinically acceptable for each outcome (delta = D). Regarding the non‐inferiority limit, in accordance with CONSORT guidelines, the authors identified the upper limit of the CI for the Test Group and the lower limit of the CI for the Control Group. These values served as the basis for calculating the maximum allowable difference. The resulting margin was graphically positioned to the right, indicating that the Ti‐Mesh Group is not considered inferior to the PTFE Group if the mean difference in quantity does not exceed 10% of clinical significance. The same procedure was performed for the variables Exposure, Infection, VBG, RR and ERR. This non‐inferiority study justified a sample size of 50 participants (allocation ratio 1/1) based on the following parameters: significance level (α) of .05; power (1 − β) of 0.9; expected proportion of 0.23 (23%) in the treatment group; expected proportion of 0.041 (4.1%) in the control group, and a margin of non‐inferiority set at 0.1. Additionally, a 10% drop rate was considered in the sample size calculation to account for potential participant attrition. The mean values for the treatment and control groups were derived from the last relevant review (Urban, Montero et al., 2019; Urban, Nagy et al., 2019).

In addition, superiority analyses were carried out to evaluate significant differences in terms of intervention group and all other variables. The comparison of means to evaluate statistically significant differences was performed by the t‐test, the Wilcoxon rank‐sum test and the Wilcoxon matched‐pairs signed‐ranks test when necessary.

For qualitative data frequencies, proportions and 95% confidence intervals for proportions were calculated. In bivariate analysis, proportions were compared using two tests. The chi‐square tests (i.e., χ²‐test) were performed when no >20% of the cells of the contingency tables had frequencies of 5 or less and that no cells had expected frequencies <1 (Cochran, 1954). If any of the observed values was <5, then a Fisher's exact test was performed. A test of normality was carried out with the Skewness/Kurtosis tests (normal distribution if p > .05). Furthermore, some variables (intervention groups and jaws) were evaluated based on the bone density (three categories) (Trisi & Rao, 1999), the pseudo‐periosteum type (three categories) (Cucchi et al., 2019) and implant stability (three categories) using the one‐way ANOVA and Tukey's post‐hoc tests. The threshold value for statistical significance was determined setting a p‐value of .05 (5%). The statistician was blinded and external to the working group. Data analysis was performed with Stata/IC software (StataCorp LLC).

3. RESULTS

3.1. Per‐protocol analysis

3.1.1. Population data

In total, 50 patients (14 males, 36 females) with a mean age of 56 years, median 59 years, within an age range of 30–70 years, were enrolled in the present study. In all, 48 patients underwent bone augmentation surgery and were treated according to the protocol described above because two patients had flap perforation at the beginning of surgery and bone augmentation was not completed.

Of the 48 patients, 29 (60.4%) were classified as ASA I, 15 (31.3%) were classified as ASA II and 4 (8.3%) were classified as ASA III, no patients with ASA IV status were treated.

42 patients were non‐smokers and six patients smoked fewer than 10 cigarettes per day; no patient smoked >10 cigarettes a day; 32 patients had no periodontal disease, whereas the 16 patients with periodontitis were treated with periodontal therapy before augmentation surgery. Of the treated defects, three were classified as small (one teeth), 31 as medium (2–3 teeth), and 14 as large (>3 teeth).

A total of 26 mandibular defects and 22 maxillary defects were treated. All treated defects were partial edentulous sites. Defect geometry was classified as ‘combined’ horizontal‐vertical in 24 cases and purely vertical in 24 cases. According to Kennedy's classification, three patients were classified as class I, 29 as class II, 7 as class III, and 9 as class IV. Additionally, scarred periosteum was present in the intervention area for 25 patients, whereas native periosteum was observed in the remaining 23 cases.

The data in per‐protocol population and in each study group, including absolute and relative frequencies, are shown in Table 1.

TABLE 1.

Demographic and clinical characteristics patients evaluated in the two study groups and in the total population, including age (mean), gender (n. and percentage), ASA classification (n. and percentage), bruxism (n. and percentage), periodontal health (n. and percentage), smoking status (n. and percentage), defect location, (n. and percentage) size (n. and percentage) and geometry (n. and percentage), Kennedy's classification (n. and percentage), and periosteum type (n. and percentage).

Parameter	PTFE group	Ti‐mesh group	Total
Age	55	57	56
Gender
M	9, 36%	5, 20%	14, 28%
F	16, 64%	20, 80%	36, 72%
ASA
1	14, 56%	15, 60%	29, 58%
2	8, 32%	7, 28%	15, 30%
3	2, 8%	2, 8%	4, 8%
Periodontitis
Yes	8, 32%	8, 32%	16, 32%
No	17, 68%	17, 68%	34, 68%
Bruxism
Yes	4, 16%	8, 32%	12, 24%
No	21, 84%	17, 68%	38, 76%
Smoke
<10 cig/die	5, 20%	1, 4%	6, 12%
No smokers	20, 80%	24, 96%	44, 88%
Defect location
Maxilla	10, 41.7%	12, 50%	22, 45.8%
Mandible	14, 58.3%	12, 50%	26, 54.2%
Defect size
Small	2, 8.3%	1, 4.2%	3, 6.2%
Medium	13, 54.2%	18, 75%	31, 64.6%
Large	9, 37.5%	5, 20.8%	14, 29.2%
Defect geometry
Purely vertical	13, 54.2%	11, 45.8%	24, 50%
Combined	11, 45.8%	13, 54.2%	24, 50%
Kennedy's class
I	3, 12.5%	2, 8.3%	3, 6.2%
II	12, 50%	15, 62.5%	29, 60.4%
III	5, 20.8%	2, 8.3%	7, 14.6%
IV	4, 16.7%	5, 20.8%	9, 18.7%
Periosteum type
Native	15, 62.5%	8, 33.3%	23, 47.9%
Scarred	9, 37.5%	16, 66.7%	25, 528%.1%

3.1.2. Surgical and healing complications

Within the entire population (n = 50), four surgical complications were observed (one in the PTFE group and three in the Ti‐mesh group), including two cases of flap perforation (one in the PTFE group and one in the Ti‐mesh group) and two related to neurological damage (in the Ti‐mesh group). Both cases of paraesthesia resolved within 3 months post‐intervention and were considered as ‘transitory’. No cases of permanent paraesthesia/dysaesthesia/hypoaesthesia/anaesthesia were recorded; then, no cases of severe bleeding or haemorrhaging were observed. Of the 50 patients, the surgical complication rates were 4% (n = 1/25) and 12% (n = 3/25) in the PTFE group and in the Ti‐mesh group (p = .297), respectively (Graph 1). The surgical complications as flap perforations were not treated: the surgeries was simply interrupted and flaps closed after, without completing the VRA procedures, whereas the neurological alterations were managed with a pharmacological therapy based on corticosteroids and neurotrophic factors (Bentelan for 14 days; Dobetin, Tiobec, and Nicetile for 3 months).

FIGURE 10.

Examples of surgical and healing complications. (a) identification of paresthesia area. (b) lingual flap perforation during regenerative surgery. (c) abscess without exposure. (d) exposure <3 mm without exudate.

The non‐inferiority analysis demonstrated the non‐inferiority of the Ti‐Mesh group compared with the PTFE group, as shown in Graph 2.

GRAPH 1.

Surgical and healing complications rates in the two study groups.

Out of 48 patients that had completed the bone augmentation surgery (T0), five healing complications occurred (three in the PTFE group and two in the Ti‐mesh group), with three arising in the early phase (within the first 3 months) and two in the late phase (after the third month).

In the PTFE group, the following healing complications were recorded: one case of early exposure <3 mm without exudate (class I) (class A), detected at the 2‐month follow‐up; one case of late exposure >3 mm without exudate (class I) (class B), identified 5 months after the regeneration procedure; one case of abscess without exposure (class IV) (class D) observed at the 1‐month follow‐up.

In the Ti‐mesh group, the healing complications included: one case of point‐like exposure (<3 mm) with exudate (class III) (class A), identified at the 1‐month follow‐up; one case of late abscess without exposure (class IV) (class D) observed at the 6‐month follow‐up.

Consequently, the healing complication rates were 12.5% (n = 3/24) and 8.3% (n = 2/24) in the PTFE group and in the Ti‐mesh group (p = .645), respectively. The non‐inferiority analysis has demonstrated non‐inferiority of the Ti‐Mesh group compared with the PTFE group in terms of healing complications, as shown in Graph 3. Examples of surgical and healing complications are shown in Figure 10.

GRAPH 2.

Error bars indicated one‐sided 95% confidence intervals of the difference in healing complication mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=0.30) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be worse than Control group in terms of surgical complication.

GRAPH 3.

Error bars indicated one‐sided 95% confidence intervals of the difference in healing complication mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=0.17) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be better than Control group in terms of healing complication.

Moreover, the exposure rates were 8.3% and 4.2% in the PTFE group and in the Ti‐mesh group (p = .561), whereas the infection rates that are the rates of complication associated with suppuration, fistula, or abscess, were 4.2% and 8.3% in the two study groups respectively (p = .561).

The non‐inferiority analysis has demonstrated non‐inferiority of the Ti‐Mesh group compared with the PTFE group in terms of exposure and infection, as shown in Graphs 4 and 5.

GRAPH 4.

Error bars indicated one‐sided 95% confidence intervals of the difference in Exposure events mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=0.17) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not include Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be better than Control group in terms of Exposure.

GRAPH 5.

Error bars indicated one‐sided 95% confidence intervals of the difference in Exsposure events mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=0.17) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be better than Control group in terms of infections.

The healing complications as exposure or abscess were treated removing the non‐resorbable device, cleaning the area with antibiotic solutions for intra‐venous administration (ceftriaxone), and closing the flaps without membranes or adjunctive devices. After the planned healing time (6–9 months), the implants were placed in association to an additional bone augmentation, if needed.

Finally, the sites associated with healing complications were divided according to the amount of graft loss: (i) complete graft loss was recorded in one case of the PTFE group and one belonging to the Ti‐mesh group; (ii) partial graft loss was observed in the remaining three cases (two in the PTFE group and one in the Ti‐mesh group); and (iii) no graft loss in any case experienced healing complication.

3.1.3. Linear measurements

The mean values of VBD were 6.08 ± 1.77 mm (range: 3.7–9.10 mm) in the PTFE group and 5.59 ± 1.68 mm (range: 3.10–9.40 mm) in the Ti‐mesh group (p = .359). At re‐entry surgery (T1), the mean values of VBG were 5.79 ± 1.71 mm (range: 3.2–8.8 mm) in the PTFE group and 5.18 ± 1.61 mm (range: 3.1–8.0 mm) in the Ti‐mesh group (p = .233). The non‐inferiority analysis has demonstrated the non‐inferiority of the Ti‐Mesh group compared with the PTFE group in terms of VBG as shown in Graph 6. The linear regeneration rates were 95.4 ± 7.6% and 92.6 ± 6.4% (p = .134), respectively. The linear measurements in each group are shown in Table 2 and in Graph 7.

GRAPH 6.

Error bars indicated one‐sided 95% confidence intervals of the difference in VBG mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=−2.7) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be worse than Control group in terms of VBG.

TABLE 2.

Linear and volumetric measurements divided by the two study groups and in the total population.

	PTFE group			Ti‐mesh group				Estimated difference	Total
	Mean/median	(SD) 95% CI	Range	Mean/median	(SD) 95% CI	Range	p‐Value	Mean 95% CI	Mean/median	(SD) 95% CI	Range
VBD (mm)	6.08/6.2	(1.77) 5.27; 6.89	3.70/9.10	5.59/5.5	(1.68) 4.85; 6.34	3.10/8.30	.36	−0.49 −1.5; 0.6	5.83/5.5	(1.72) 5.30; 6.36	3.10/9.10
VBG (mm)	5.80/5.3	(1.71) 5.01; 6.58	3.20/8.8	5.18/5.15	(1.61) 4.47; 5.89	3.10/8	.23	−0.6 −1.6; 0.4	5.48/5.3	(1.67) 4.97; 5.99	2.9/8.8
PBV (cc)	1.49/1.47	(0.51) 1.25; 1.72	0.54/2.59	1.42/1.3	(0.59) 1.16; 1.69	0.33/3.21	.70	−0.94 −0.4; 0.3	1.46/1.40	(0.55) 1.29; 1.62	0.33/3.20
ABV (cc)	1.62/1.58	(0.48) 1.40;1.84	0.86/2.48	1.42/1.30	(0.58) 1.17; 1.68	0.35/3.10	.22	−0.2 −0.5; 0.3	1.52/1.42	(0.54) 1.35; 1.68	0.35/3.1
RBV (cc)	1.46/1.48	(0.48) 1.24; 1.68	0.52/2.25	1.26/1.17	(0.55) 1.01; 1.50	0.25/2.71	.20	−0.2 −0.5; 0.1	1.35/1.25	(0.52) 1.19; 1.52	0.25/2.71
LBV (cc)	0.16/0.16	(0.1) 0.13; 0.21	0.04/ 0.34	0.17/0.14	(0.1) 0.12; 0.21	0.05/0.39	.87	0.002 −0.05; 0.06	0.17/0.15	(0.09) 0.13; 0.19	0.04/0.39
RR (%)	99.5/99.5	(14.44) 92.93; 106.07	44.74/132.79	87.0/86.40	(8.93) 83.01; 90.03	63.93/111.84	.013	−12.5 −19.9; −5.2	93.09/90.72	(13.39) 88.96; 97.21	44.74/132.79
ERR (%)	88.9/92.0	(9.15) 84.72; 93.05	60.47/100	86.7/88.66	(6.33) 83.94; 89.55	60.47/100	.37	−2.1 −7; 2.7	87.79/89.47	(7.82) 85.38; 90.19	60.47/100

Note: Mean, median, standard deviation (SD), confidence interval (95% CI), range, p‐value and estimated difference between the two groups (mean and 95% CI).

Abbreviations: ABV, augmented bone volume; ERR%, effective regeneration rate; LBV, lacking bone volume; PBV, planned bone volume; RBV, regenerated bone volume; RR%, regeneration rate; VBD, vertical bone defect; VBG, vertical bone gain.

GRAPH 7.

Linear measurements in the two study groups. VBD, vertical bone defect; VBG, vertical bone gain.

3.1.4. Volumetric measurements

The per‐protocol population included all 43 patients who had a successful bone augmentation surgery (T0) and were re‐evaluated after 6–9 months for the implant surgery (T1).

During digital workflow and volumetric analyses, the pre‐operative CBCT revealed mean values of PBV 1.49 ± 0.51 cc in the PTFE group (range: 0.54–2.59 cc) and 1.42 ± 0.59 cc (range: 0.33–3.2 cc) in the Ti‐mesh group.

At the re‐entry surgery (T1), the post‐operative CBCT revealed mean values of ABV 1.62 ± 0.48 cc (range: 0.86–2.48 cc) and 1.42 ± 0.58 cc, (range: 0.35–3.1 cc), respectively (p = .223), whereas the values of RBV were 1.46 ± 0.48 cc (range: 0.52–2.25 cc) in the PTFE group and 1.26 ± 0.55 cc (range: 0.25–2.71 cc) in the Ti‐mesh group (p = .204). Regarding LBV, the mean values were similar in both study groups: 0.16 ± 0.1 and 0.17 ± 0.1 cc, respectively (p = .866). Volumetric measurements in the per‐protocol population are given in Table 2; and relative percentages of ABV, RBV, and LBV are shown in Graph 8.

GRAPH 8.

Rates of bone volumes in the two study groups. ABV, RBV + LBV: Total area; LBV, lacking bone volume; RBV, reconstructed bone volume.

The regeneration rates (RR%) were 99.5% and 87.0% in the PTFE group and in the Ti‐mesh group, demonstrating a statistically significant difference (p = .013). Also the effective regeneration rate (ERR%) was higher in the PTFE group (88.9%) than in the Ti‐mesh group (86.7%), but statistical analysis did not show any significant differences (p = .376). The non‐inferiority analysis has demonstrated the non‐inferiority of the Ti‐Mesh group compared with the PTFE group in terms of RR% and ERR%, as shown in Graphs 9 and 10. The regeneration rates in the per‐protocol population are given in Table 2 and Graph 11.

GRAPH 9.

Error bars indicated one‐sided 95% confidence intervals of the difference in RR mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=−2.7) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be worse than Control group in terms of RR%.

GRAPH 10.

Error bars indicated one‐sided 95% confidence intervals of the difference in ERR mean values between the Test and Control groups (Mesh minus e‐PTFE). The red broken line delineating the difference in the score (=−2.7) shows the non‐inferiority margin (delta); Tinted area indicates zone of non‐inferiority. The CI does not includes Δ and the data prove non‐inferiority of Ti‐mesh group compared with e‐PTFE group. Although there is no statistically significant difference between the two treatments, Test group tends to be worse than Control group in terms of ERR%.

GRAPH 11.

Regeneration rates in the two study groups. ERR%, effective regeneration rate; RR%, regeneration rate. *Statistically significant difference (p < .05).

In all the cases of healing complications (exposure of the membrane or infection), the mean RR% was 52.6 ± 0.19%, specifically 54.5 ± 22.6% in the PTFE group and 49.6 ± 20.2% in the Ti‐mesh group (p = .800).

In the intent‐to‐treat population, including all cases of partial graft loss and excluding the two cases of total graft loss (46 patients), the regeneration rates (RR%) were 96.3 ± 18.3% in the PTFE group and 85.9 ± 9.9% in the Ti‐mesh group (p = .0214) and the effective regeneration rates (ERR%) were 89.8 ± 9.3% in the PTFE group and 87.32 ± 6.8% in the Ti‐mesh group (p = .0770).

3.1.5. Clinical data

In the PTFE group at re‐entry surgery (T1), type‐1 pseudo‐periosteum was found in 15 patients (71.4%) 8 in mandible and 7 in maxilla; type‐2 in 5 patients (23.8%) 3 in mandible and 2 in maxilla; and type‐3 in only one patient (4.8%) in mandible. In the Ti‐mesh group, type‐1 was observed in 10 patients (45.5%) 4 in mandible and 6 in maxilla; type‐2 in 7 patients (31.8%) 5 in mandible and 2 in maxilla; and type‐3 in 5 patients (22.7%) 3 in mandible and 2 in maxilla. These results were not statistically significant. The pseudoperiosteum type distribution between the two groups is shown in Graph 12 and Table 3.

GRAPH 12.

Distribution rates of periosteum type, bone density, and implant stability comparing the two study groups.

TABLE 3.

Frequency table relating the pseudoperiosteum type (I, II, III) and the two study groups.

Study group	Pseudoperiosteum
	I	II	III	Total
PTFE‐group	15	5	1	21
	71.4%	23.81%	4.8	100%
	60%	41.7%	16.7%	48.8%
Ti‐mesh group	10	7	5	22
	45.4%	31.8%	22.7%	100%
	40%	58.3%	83.3%	51.2%
Total	25	12	6	43
	58.1%	27.9%	13.9%	100%
	100%	100%	100%	100%

In the PTFE group, hard bone density was found in 4 patients (all in mandible); medium density in 9 patients (5 in mandible and 4 in maxilla); and soft in 8 patients (3 in mandible and 5 in maxilla). In the Ti‐mesh group, hard bone density was recorded in 11 patients (10 in mandible and 1 in maxilla); medium density in 7 patients (2 in mandible and 5 in maxilla); and soft in 4 patients (all in maxilla). No statistically significant differences were observed between the two study groups. The bone density distribution between the two groups is shown in Table 4.

TABLE 4.

Frequency table relating the bone density type (hard, medium, soft) and the two study groups.

Study group	Density
	Hard	Medium	Soft	Total
PTFE‐group	4	9	8	21
	19%	42.9%	38.1	100%
	26.7%	56.2%	66.7%	48.8%
Ti‐mesh group	11	7	4	22
	50%	31.8%	18.8%	100%
	73.3%	43.7%	33.3%	51.2%
Total	15	16	12	43
	34.9%	37.2%	27.9%	100%
	100%	100%	100%	100%

In total, 115 implants were placed using computer‐guided surgery in the augmented sites, without any need for additional bone augmentation. Fifty‐five implants were placed in patients belonging to the PTFE group (56.4% in maxilla and 43.6% in mandible); implant stability was assessed as high (>35 N/cm) in 66.7% of implants, medium (between 15 and 35 N/cm) in 28.6%, and low (<15 N/cm) in 4.8%. Sixty implants were inserted in the Ti‐mesh group (48.3% in maxilla and 51.7% in mandible); implant stability was defined as high in 59.1% of the sites, medium in 36.4%, and low in 4.5%. No significant differences were observed between the two study groups (p = 1.00).

Implant stability distribution between the two groups is given in Table 5.

TABLE 5.

Frequency table relating the implant stability (high, medium, low) and the two study groups.

Study group	Implant stability
	High	Medium	Low	Total
PTFE‐group	14	6	1	21
	66.7%	28.6%	4.8%	100%
	51.8%	42.9%	50%	48.8%
Ti‐mesh group	13	8	1	22
	59.1%	36.4%	4.5%	100%
	48.1%	57.1%	50%	51.2%
Total	27	14	2	43
	62.8%	32.6%	4.6%	100%
	100%	100%	100%	100%

Considering the total number of inserted implants, implant stability was found to be high in 36.8%, medium in 52.6% and low in 10.5% of implants placed in maxilla, whereas it was high in 83.3%, medium in 16.7% and low in 0% of implants placed in mandible, with a statistically significant difference (p = .005). Implant stability distribution between maxilla and mandible is shown in Table 6. In terms of implant stability between maxilla and mandible within the two groups, no statistically significant difference was observed (p = .869 for maxilla and p = .590 for mandible), the corresponding data are contained in Tables 7 and 8.

TABLE 6.

Frequency table relating the implant stability (high, medium, low) and the jaw.

Jaw	Implant stability
	High	Medium	Low	Total
Superior	7	10	2	19
	36.8%	52.6%	10.5%	100%
	25.9%	71.4%	100%	51.2%
Inferior	20	4	0	24
	83.3%	16.7%	0%	100%
	74.1%	28. 6%	0%	55.8%
Total	27	14	2	43
	62.8%	32.6%	4.6%	100%
	100%	100%	100%	100%

TABLE 7.

Frequency table relating the implant stability in maxilla (high, medium, low) and the two study groups.

Study group	Implant stability in maxilla
	High	Medium	Low	Total
PTFE‐group	3	5	1	9
	33.3%	55.6%	11.1%	100%
	42.9%	50%	50%	48.8%
Ti‐mesh group	4	5	1	22
	40%	50%	10%	100%
	57.1%	50%	50%	51.2%
Total	7	10	2	19
	36.8%	52.6%	10.5%	100%
	100%	100%	100%	100%

TABLE 8.

Frequency table relating the implant stability in mandible (high, medium, low) and the two study groups.

Study group	High	Medium	Total
PTFE‐group	9	3	12
	75%	25%	100%
	45%	75%	50%
Ti‐mesh group	11	1	12
	91.7%	8.3%	100%
	55%	25%	50%
Total	20	4	24
	83.3%	16.7%	100%
	100%	100%	100%

At the re‐opening surgery (T2), implant osseointegration in the 2 groups was 98.2% and 98.3% respectively, due to one implant failure in each group (one implant failed in the maxilla in PTFE group; and one in the mandible in Ti‐mesh group).

Distribution and relative percentages of pseudoperiosteum type, bone density, and implant stability are given in Graph 6.

In per‐protocol population, 107 implants were inserted into 42 patients: 49 implants in PTFE group and 58 implants in Ti‐Mesh group. All implants were restored using a restoration screw‐retained on multi‐unit abutments or directly on the implants, depending on the clinician's choice.

The mean marginal bone level at baseline was 0.55 ± 0.43 mm, with statistically similar values in PTFE group (0.55 ± 0.54 mm) and in the Ti‐mesh group (0.55 ± 0.33 mm).

After 1 year of follow‐up, the mean peri‐implant bone loss was 0.23 ± 0.19 mm, with no significant differences between PTFE group (0.25 ± 0.21 mm) and Ti‐mesh group (0.20 ± 0.18 mm).

All osseointegrated implants were still functioning and successful, giving an implant success rate of 100% in both groups.

4. DISCUSSION

Guided bone regeneration (GBR) is a well‐established technique in dental implantology aimed at improving alveolar bone defects. This technique involves the use of non‐resorbable d‐PTFE membranes or titanium meshes as barriers to achieve an effective vertical ridge augmentation (Elnayef et al., 2017; Retzepi & Donos, 2010; Urban, Montero et al., 2019; Urban, Nagy et al., 2019). Despite its success, traditional GBR techniques have been associated with high complication rates, including early or late exposure of the barrier, can lead to infections, and are often dependent on the operator's skills and experience (Alotaibi et al., 2023; Gallo & Díaz‐Báez, 2019; Lizio et al., 2022; Poli et al., 2022; Vroom et al., 2022).

Over the years, new devices such as customized titanium meshes and reinforced PTFE meshes and new digital approaches for GBR have been developed to facilitate the surgical procedures, to reduce the risk of complications, and to improve bone regeneration (Chiapasco et al., 2021; Cucchi et al., 2022; Urban et al., 2021).

In the scientific literature, there are no randomized or non‐randomized clinical studies comparing these new digital approaches and new devices for GBR: the present study aimed at comparing the effectiveness of digital customized CAD/CAM meshes and reinforced d‐PTFE meshes with digital approach. The primary outcome was the surgical and healing complication rates, whereas vertical bone gain and regeneration rates based on volumetric changes were secondary outcomes.

To compare these parameters, the study used a non‐inferiority design due to the lack of prior comparative studies on these novel devices, whereas the sample size was calculated on the basis of results derived from the last relevant review (Urban, Montero et al., 2019; Urban, Nagy et al., 2019): 25 patients were enrolled and treated for each study group. No other randomized clinical trials about VRA involved such a high number of participants.

The rationale to use the healing complication rates as primary outcome is due to the fact that all measures of treatment efficacy (e.g., vertical bone gain, regenerated bone volume, or regeneration rates) are consequent upon and related to uneventful healing. If a healing complication occurs, all variables are negatively influenced: as reported in this study, every healing complication was related to a partial or complete bone graft loss. Consequently, the authors consider the absence of healing complication the primary end‐point in every bone augmentation to achieve success.

In a recent literature review, an average of 16% complications was reported across 32 studies, 761 participants, and 943 defects, for the management of alveolar ridge defects with different techniques: onlay graft, inlay graft, d‐PTFE, e‐PTFE, titanium meshes, resorbable membranes, distraction osteogenesis. The network metanalysis suggested that GBR performed best among the vertical augmentation techniques with regard to healing complication with d‐PTFE performing slightly better than e‐PTFE (Alotaibi et al., 2023). Operator experience is directly linked to the incidence of complications.

In the present study, the percentages of surgical complications in both groups (4% vs. 12%) were less than those reported in the above‐mentioned systematic review (17%). The surgical complications, including two cases of flap perforation and two related to transient neurological damages, highlight the importance of very careful surgical techniques that need to take in consideration every detail and the need for precise and comprehensive information given to the patient prior to obtaining consent.

The flap perforations could be explained by the fact that all the surgeries were conducted with video recording during live surgery, which complicated the execution of surgical procedures. There may have been added pressures or distractions that made it more difficult to perform the procedures without accidentally damaging the surgical flaps. Then, it is well known that VRA procedures are challenging and operator‐dependent and the number of the surgeon's previous surgeries (<200 procedures) could be directly linked to the incidence of complications. However, no permanent paraesthesia/dysaesthesia but only two transient paraesthesia were counted in this study population, that solved spontaneously within 3 months.

Similarly, the percentages of healing complications in the two study groups (12% vs. 8%) were less than those reported in this updated review (16%). The healing complications observed in the present study were not so high considering the surgeon's experience (about 200 VRA procedures) and the severity of bone defects (mean values >5 mm). The present investigation about the exposure rates and infection rates in the two study groups showed that no statistically significant differences were found with the GBR technique regarding surgical and healing complications with the use of d‐PTFE meshes and titanium meshes. Moreover, the low healing complication rates, especially in the upper jaw compared with lower jaws (Cucchi et al., 2017; Poli et al., 2022), can be explained by the adoption of the Bichat's buccal fat pad technique above GBR. As demonstrated, the use of the buccal fat pad is an effective technique in ensuring primary intention healing, due to the pluripotent cells it contains and the tissue thickening it creates during wound healing (Cucchi et al., 2023).

As to customized CAD/CAM titanium meshes, the most recent articles reported an exposure rate ranging from 21% (Chiapasco et al., 2021) to 53% (Lizio et al., 2022) and these values were higher than those observed in the present study, where only one device showed an early exposure (4%). These differences could be explained by the type of mesh with greater thickness or unfavourable design (Lizio et al., 2022) or by the type of treated sites with greater extent or more severity (Chiapasco et al., 2021). Moreover, the fact that several operators participated in those studies could be another reason for the high rate of exposure, as opposed to the present study where one operator performed all the surgeries. In regard to reinforced PTFE meshes, the only article published by Urban et al. (2021) showed a complication rate of 3% that is significantly less than those reported in the present study. This could be explained by the surgeon's greater experience and the similarities of the traditional approach with ePTFE or dPTFE membranes.

A total graft loss occurred only in two patients (one per group), in the remaining 3 patients in whom the healing complication resulted in partial graft loss, implants were nevertheless placed with additional GBR with resorbable membranes. These findings were similar to those reported by other authors, where implant treatment was pursued in most of the cases with exposure and partial bone loss (Chiapasco et al., 2021; Cucchi et al., 2021; Hartmann et al., 2022; Lizio et al., 2022).

The mean vertical bone defect in the present study was about 6 mm in the PTFE group and about 5.5 mm in the Ti‐mesh group, and the vertical bone gain measured as linear regeneration rate after bone augmentation was about 95% and 93%, respectively. Even though the PTFE group had higher values than the Ti‐mesh group, no significant differences were observed according to the device used.

These values were similar to those reported in other studies. In a case series study, Urban et al. (2021) reported a mean VBD of 5.5 mm and a linear regeneration rate of about 97% using reinforced PTFE meshes filled with a 50:50 mixture of an organic bovine bone and autogenous bone. The authors had a complete regeneration in 89% of cases: this healing pattern was observed in all sites with initial deficiencies <5 mm, in 96% of sites with deficiencies ranging from 5 to 8 mm, and in 89% of sites with deficiencies exceeding 8 mm. Similarly, using customized CAD‐CAM titanium meshes, in a retrospective clinical study, Sagheb et al. (2017) obtained a VBG of about 6.5 mm and Chiapasco et al. (2021) observed a VBG of about 5 mm using customized CAD/CAM meshes. Since no information about initial bone defects was reported, linear regeneration rates cannot be calculated in those studies.

The recent innovations in software for 3D analysis made it possible to measure the volume of regeneration to compare the planned bone volume and the regenerated bone volume, obtaining the so‐called regeneration rates (Cucchi et al., 2021). These values can give information about the effectiveness of the VRA procedures to have a better comparison between different techniques.

The mean volumes of augmentation were 1.6 and 1.4 cc for the PTFE and Ti‐mesh groups, respectively, with no differences between them. Also, these values were similar to those reported by other authors, ranging from 0.8 to 1.5 cc (Chiapasco et al., 2021; Cucchi et al., 2021; Lizio et al., 2022).

In this study, both groups had sites with volumes of augmentation up to 2.5 cc, confirming that both surgical techniques were effective in treating severe defects, as they achieved RR >90%.

The percentages of RR were 99.5% in the PTFE group and 87.0% in the Ti‐mesh group, if the planned bone volume was used as the reference. However, the authors introduce the ERR because in the PTFE group, despite the 3D planning and the use of mesh replica, the augmented volumes (ABV) were higher than the planned volumes (PBV) in the most of sites, with significant difference. This is due to the elasticity and flexibility of the non‐resorbable device, which permitted an additional increase in volume after fixation due to the compaction of the graft material.

So it is worth noting that while RR and ERR were similar in the Ti‐mesh group, different values were reported in the PTFE group. This difference was related to the mechanical features of the device used: the rigidity of titanium meshes did not permit any extension, while PTFE membranes were able to expand when filled with the graft. The expansion of PTFE membranes caused the increase of ABV values compared with PBV values, thus reducing the ERR (RBV/ABV) with respect to RR (RBV/PBV).

Although the surgical technique based on non‐resorbable membrane allowed the clinicians to increase the augmentation volume, resulting in higher values of RR compared with customized CAD/CAM titanium meshes, the percentages of ERR were similar in both study groups (89% vs. 87%) confirming that the barrier device could not influence the regeneration in terms of bone formation and remodelling, that appear more influenced by the site's potential rather than the device material.

The RR values were comparable to those reported in a recent study by the same authors, comparing customized Ti‐mesh alone or covered with resorbable membrane that showed rates of 74% and 82%, respectively (Cucchi et al., 2021).

When discussing the regeneration rate, it is interesting to note that in cases of healing complications, a mean RR value of 53% was obtained, and that is very similar than the 64% reported in the previous study. It is important to underline that not all healing complications led to total bone loss and in most cases implant placement can be obtained with or without additional GBR, as reported by many authors (Chiapasco et al., 2021; Hartmann & Seiler, 2020; Li et al., 2021; Lizio et al., 2022).

In regard to pseudo‐periosteum, that represents the missing bone volume, the statistical analysis did not reveal any significant differences, but some trend could be identified. In the PTFE group, where a long‐lasting resorbable membrane was used over the reinforced PTFE mesh, at re‐entry surgery in most of the cases (71%) pseudo‐periosteum was clinically assessed as type I, whereas in the Ti‐mesh group, where a short‐lasting resorbable membrane was used over the titanium mesh, in most of the cases (55%) type II and type III were found. In accordance with these results, the titanium mesh covered with a long‐lasting membrane had pseudo‐periosteum of type I in most of the cases (67%), whereas titanium mesh alone led to type II and type III in most of the cases (53%) (Cucchi et al., 2021). The slow resorption of resorbable membranes over reinforced PTFE or titanium meshes increased the barrier effect increasing the bone regeneration and reducing the formation of pseudo‐periosteum. However, different variables can influence pseudo‐periosteum formation and they should be investigated in further studies. Other authors suggested using a double layer of resorbable membrane over titanium meshes to reduce the layer of pseudo‐periosteum, obtaining type I in most of the cases (69%) (Li et al., 2021).

Regarding bone density, no statistically significant differences between the groups were observed, but some trends should be underlined. In the PTFE group, hard bone density was found in 17% of the sites, whereas in the Ti‐mesh group hard bone density was recorded in 46% of the sites. Then, it is interesting to note that independently of the device used, hard bone was only found in the mandible in all cases except one. A direct correlation between the densities of the pre‐existing and reconstructed bone has been first demonstrated using PTFE membrane (Simion et al., 1998) or titanium mesh (Corinaldesi et al., 2007) for vertical ridge augmentation in comparative histologic studies.

To compensate the different bone density and bone remodelling in the upper jaw with respect to the lower jaw, different healing times were adopted: 6 months in the mandible and 9 months in the maxilla before device removal and implant placement.

A total of 105 implants were placed in augmented sites without differences with regard to implant stability. The most of implants showed stability >15 N/cm, and only 5% of the implants had stability of <15 N/cm. Even if unsupported by histological data, from a clinical standpoint these observations indicated a good quality of the regenerated bone, that was also confirmed by the osseointegration rates of 98% in both groups.

The main limitations of this study are the short‐term follow‐up and the number of patients, that should be higher to be able to draw strong conclusions about some variables such as RR or ERR and pseudo‐periosteum or density of regenerated bone. Then, subjective measurements of RBV and LBV could be considered a limit of this study.

Otherwise, the study undoubtedly showed many strengths: first of all, the randomized and blinded design of the study protocol, the 3D virtual planning and digital approach of all treated sites, the customization of the barrier device, the analysis of both linear and volumetric measurements to evaluate vertical bone gain and regeneration rates, and finally the completeness of data.

In conclusion, this study showed the potential of digital approaches in GBR in overcoming complications and achieving significant vertical bone augmentation. These innovative approaches, utilizing customized CAD/CAM titanium meshes or digitally‐designed reinforced d‐PTFE meshes, offer promising results in terms of reduced complications, improved bone regeneration, and successful implants, without significant differences between the two surgical techniques.

5. CONCLUSIONS

This randomized clinical trial confirmed that both digital surgical approaches are reliable and effective to perform the vertical ridge augmentation in patients with localized defects (>5.5 mm).

The customized CAD/CAM titanium meshes seem to be not inferior to reinforced PTFE meshes in terms of surgical complications (4% vs. 12%) and healing complications (12.5% vs. 8.3%).

Although PTFE meshes showed higher vertical bone gain (5.8 mm vs. 5.2 mm), regeneration rate (99.5% vs. 87%), and effective regeneration rate (88.9% vs. 86.7%) than Ti‐meshes, no statistical differences were found for all these variables, and the statistical analysis confirmed the equivalence of PTFE meshes and Ti‐meshes. Further studies should be carried out to investigate more deeply these issues.

Similarly, no significant differences were observed for pseudo‐periosteum type and bone density even if PTFE mesh provided the most cases of type‐1 pseudo‐periosteum (71.4% vs. 45.4%) and Ti‐mesh the most cases of hard bone density (19% vs. 50%).

Finally, high implant stability (66.7% vs. 59.1%) and high osseointegration rates (98.2% vs. 98.3%) were similar in the two‐study group, that appeared equal each other.

AUTHOR CONTRIBUTIONS

Alessandro Cucchi: Conceptualization (lead); investigation (lead); Resources (equal); Validation (supporting). Sofia Bettini: Data curation (lead); Methodology (equal); Writing – original draft (lead); Resources (supporting). Lucia Tedeschi: Data curation (supporting); Writing – original draft (equal); Software (equal). Istvan Urban: Writing – review and editing (equal). Debora Franceschi: Reviewing and editing (equal). Antonino Fiorino: Data curation (equal); Formal analysis (lead); Methodology (supporting); Software (lead). Giuseppe Corinaldesi: Methodology (lead); Project administration (lead); Resources (equal); Supervision (lead).

CONFLICT OF INTEREST STATEMENT

The authors directly involved in the study protocol report no conflicts of interest related to this study.

Supporting information

Data S1:

Cucchi, A. , Bettini, S. , Tedeschi, L. , Urban, I. , Franceschi, D. , Fiorino, A. , & Corinaldesi, G. (2024). Complication, vertical bone gain, volumetric changes after vertical ridge augmentation using customized reinforced PTFE mesh or Ti‐mesh. A non‐inferiority randomized clinical trial. Clinical Oral Implants Research, 35, 1616–1639. 10.1111/clr.14350

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.