The inclusion criteria were as follows:

• vertical insufficiency that would result in poor crown-to-implant ratio
• edentulous regions with a height of ≤7 mm and a width of ≤4 mm
• combined horizontal with vertical defects
• capacity to understand and accept the written conditions of the study.

The exclusion criteria were as follows:

• insufficient oral hygiene
• smoking habit
• acute local and systemic infections
• uncontrolled systemic disease
• patients who underwent radiotherapy
• patients, undergoing immunosuppressive therapy or immunocompromised patients.

After standard patient history was taken, the patients were placed in either group with the use of a simple randomization technique, carried out by a coin toss. The study subjects were enrolled, planned, and treated from 2021 to 2023: 20 patients were assigned to the test group (Ti mesh) and 20 patients to the control group (Ti d-PTFE). Each patient received written information and provided written consent before any study-related procedure.

All candidates for the study underwent a CBCT procedure, where the baseline data for the available bone height and width was gathered.

The DICOM data of the subjects that were designated to the test group was sent to a company that has division for designing and manufacturing patient specific titanium printed medical devices (Biotec Srl). The company then transferred the DICOM files obtained from the CBCT in stereolithography (STL) format (Fig. 1).

From there on, the insufficient bony tissues were recreated virtually, and a mesh-like structure was designed over the simulated “regenerated tissue” (Fig. 2). In addition to that, the openings for the placement of the fixation screws were located away from any vital structures to avoid their impairment. When the digital process was finished, the firm contacted us and asked if any changes needed to be made. The process ended with us agreeing (in writing) to the digital design of the mesh, which gave our partners a green light regarding the manufacturing process which consisted of 3D printing the mesh through the process of select laser sintering (Fig. 3).

From there on, the patients from the test and control group were scheduled for the identical surgical procedure.

Figure 1.

Stereolithographical model of an atrophic alveolar ridge.

Figure 2.

Regenerated tissue over atrophic ridge and design of the patient specific mesh.

Figure 3.

Printed patient specific titanium meshes.

Written consent was obtained from all patients before the surgery. After anesthesia, the procedure began with a midcrestal incision in keratinized tissue, which was extended buccally two teeth away (Fig. 4a) and lingually (Fig. 4b) one tooth away from the edentulous site creating the so called “asymmetrical flap”.^[6] From there on, vertical releasing incisions were created alongside the line angles of the teeth involved in the sulcular incision.

Since one of the principles for successful guided bone regeneration is primary intention closure, the flaps needed to be mobilized to facilitate the soft tissue closure over the additionally placed bony tissue. The buccal flap was extended with a simple periosteal releasing incision. In the mandibular cases the lingual flap’s length was managed according to the varying anatomical zones, composing its structure (Fig. 5).

The lingual flap is divided in three zones as proposed by Urban et al.^[7]: the first zone is the retromolar area (Fig. 6a) from which the soft tissue is only reflected with a periosteal elevator; the second zone involves the mylohyoid area (Fig. 6b), and its management involves blunt dissection of the superficial fibers of the muscle; and lastly the third zone – premolar region (Fig. 6c) – in which the tissues are mobilized by a simple periosteal releasing incision.

The aftermath of this intricate lingual flap management technique is a three-fold increase of its length (Fig. 7).

The rest of the surgery involved, perforating the cortical plate of the atrophic alveolar ridge to facilitate angiogenesis and harvesting autogenous bone from the oblique line of the mandible with a scraper and mixing it with a xenograft in a ratio of 1:1. This bony mixture was placed either on the inner surface of the mesh in the case of the patients from the test group (Fig. 8a) or on the atrophic ridge in the cases involving Ti d-PTFE membrane (Open Tex-TR, Purgo) (Fig. 8b). After that the meshes/membranes were stabilized with titanium screws (Myungsung CNM).

The surgical procedure ends with a bilayered closure of the augmented region with PTFE sutures – horizontal mattress approximately 8 mm to 10 mm away from the wound edges and coronal to them – simple interrupted sutures (Biotex, Purgo). The patients were given postoperative instructions and were prescribed antibiotics (augmentin 100 mg) and NSAIDs (nimesil 100 mg). Sutures were left undisturbed for 2–3 weeks and the augmented bone was left to heal for at least 6 months. At re-entry, flaps were elevated, the screws and membranes/meshes were removed and implants (Megagen; Anyridge implants) were placed in the healed sites (Fig. 9).

Figure 4.

Midcrestal and sulcular incision: a) buccal view; b) lingual view.

Figure 5.

Initial length of lingual flap.

Figure 6.

Lingual flap management: a) retromolar region; b) mylohyoid part; c) premolar area.

Figure 7.

Lingual flap length after the application of Urban’s technique.

Figure 8.

Placement of bony tissue: a) in titanium mesh; b) on top of atrophic ridge with Ti-PTFE membrane, supporting on the posterior.

Figure 9.

Placement of implants in augmented ridge.

Alveolar ridge height and width were recorded before the surgical procedure. Adjacent anatomical structures were used as reference points to position the coronal slices of the CBCT. After that the values were cumulated separately and a mean baseline height and width were documented for every patient. After the healing period, the same process was repeated, using the same anatomical landmarks and post-procedure mean height and width were recorded and later compared quantitively (Fig. 10). As for the healing complications, they were assessed qualitatively using the classification by Fontana et al.^[8]

Figure 10.

Pre- and post-op measurement of height and width.

We used the Minitab statistical software (2022) to estimate the power of a two-sample t-test with a sample size of 20 in both study groups. We entered the following parameters: sample size =20, type I error =0.05, assumed minimal difference =2.5, and assumed standard deviation =2.6. The calculation yielded a power of 0.893, indicating almost a 90% accuracy of correctly rejecting a true null hypothesis. The data were analyzed using SPSS v. 27 (2020). Continuously measured variables were screened for normality using the Shapiro-Wilk test. In the presence of normal distributions, the mean values and standard deviations (SDs) were reported, and between-group comparisons were carried out with the independent samples t-test. Non-normally distributed variables were described with median values and interquartile ranges (IQRs), and the Mann-Whitney U test was used to compare the groups. The categorical data were presented as numbers and percentages (%), and relationships were determined through the chi-square test and Fisher’s exact test. All statistical tests were two-tailed and performed at a type I error (α) of 0.05.

The alveolar ridge height values, including baseline, acquired, and absolute gain, were normally distributed in both groups, and the results were illustrated through individual value plots in Fig. 11. There was no significant difference (p=0.567) in the mean baseline height between the Ti-mesh group (9.21±2.67 mm) and the control group (9.82±3.29 mm). The Ti-mesh group’s acquired height was 12.86±3.14 mm versus 14.07±3.14 mm in the control group (p=0.295). There was no significant difference in the absolute bone gain (acquired – baseline) for the Ti-mesh group (3.65±1.73 mm) and that for the control group (4.24±2.19 mm) (p=0.400).

Figure 11.

Linear changes in the vertical dimension.

While the alveolar ridge width values for the baseline and acquired measures were not normally distributed, the absolute width gain had a normal distribution in both groups. Because of this, box plots were used to show the findings for the first two variables (Fig. 12a) and individual value plots for the absolute width gain (Fig. 12b). The median baseline width for the Ti-mesh group was 3.07 mm (IQR=1.15) compared to 2.88 mm (IQR=0.77) for the control group (p=0.580). The acquired width had a median of 5.79 mm (IQR=0.33) in the Ti-mesh group and a median of 5.72 mm (IQR=0.54) in the control group. The mean absolute bone gain (acquired – baseline) for the test group amounted to 2.48±1.03 mm and to 2.60±0.82 mm for the control group (p=0.719).

Figure 12.

Linear changes in the horizontal dimension

The overall complications rate in the two groups is illustrated in Fig. 13a, and the distribution of the complications according to Fontana’s classifications is given in Fig. 13b.

Figure 13.

a) Overall complications rate; b) Complications rate according to Fontana classification.

The differences in our results, regarding vertical and horizontal absolute bone gain in the two groups, were small and statistically insignificant. The bone gain, recorded by our group was comparable with the results of other authors using custom printed titanium meshes^[9–12] or Ti d-PTFE membranes^[13–15]. The literature also shows several studies that surpass the bone gain achieved by us in our study with 3D printed meshes^[16–19] or with Ti d-PTFE membranes^{[20, 21]}.

An interesting observation can be made with regards to the distribution of the results from our test and control groups. In our control group, the linear results for vertical and horizontal gain appear in a wider range in comparison with the mean value, whereas the exact opposite observation can be made for the regeneration parameters from the subjects of the test group. It could be hypothesized that the use of a graft stabilization device such as a custom designed titanium mesh leads to more predictable outcomes due to its “built-in” vertical and horizontal parameters. With that being said, the larger diameter of the openings on the surface of the titanium lattice structure did not show statistically significant increase of the regenerated tissue due to “dual vascularization” (bone marrow and periosteal blood vessels).

However, when the correlation of the linear results from different studies regarding GBR is examined, it should always be taken with a grain of salt. The gathered data from systemic reviews on the subject^{[22, 23]} show immense heterogenicity – every author has a different understanding of a baseline and acquired bone height and width, and the compared measurements were done both clinically and radiographically. A recommendation for future research could be finding or developing an objective and universal measurement method for the documentation and juxtaposition of linear regenerative bone changes.

Healing complications after GBR procedures are not an uncommon event. This is mostly because the soft tissues must cover the additional bone volume and the tension in the suture line could be a reason for a wound dehiscence and subsequent graft destabilization. Our study unfortunately did not lack post-op complications which proved to be detrimental to the outcome of the regenerative procedure. However, the complication rate between the two group showed no statistical significance and the custom fit of the mesh in the test group failed to show superiority in the inquiry for a better graft stabilization device. Quantitively, our results with regards to complications in the test group^{[19, 24]} and in the control group^{[14, 25]} were comparable to those of other studies.

Another interesting observation with regards to the clinical behavior of the mesh/membrane, and the related ph-mediated resorption of the graft, could be made. The results of our study concluded that Ti d-PTFE performed better in the case of an early exposition of the membrane. This could be because their microporosity maintains the integrity of the bone graft, at least to some extent. Contrary to that, the mesh structure does not provide any protection for the augmented volume at all. When late exposures were examined however, the roles were reversed. The fact that a thick protective layer of “pseudoperiosteum” (largely composed of natural connective tissue) was formed over the graft surface in the subjects from the test group led to less infection in the dehisced sites. In comparison, the exposed synthetic PTFE led to acute infection of the adjacent soft tissues sooner or later.