Marginal gap of three-dimensional printed full-arch frameworks supported by all-on-four and all-on-six implant designs

Abstract

Aim:

The aim of this study was to evaluate the marginal gap of full-arch frameworks (FAFs) supported by all-on-four and all-on-six implant designs, fabricated using different manufacturing technologies.

Settings and Design:

This was an in vitro study.

Materials and Methods:

Fifteen titanium FAFs were fabricated using milling and three-dimensional printing techniques: selective laser melting (SLM) and electron beam melting (EBM) (n = 5/group). The marginal gap between the framework and abutment was measured using a microscope with 1 μm accuracy. Measurements were taken three times by a calibrated examiner (intraclass correlation coefficient of 0.996; P < 0.001) at the buccal and lingual interface between the abutment and the framework.

Statistical Analysis Used:

A two-way ANOVA was applied to assess the effects of implant design and manufacturing technology (α = 0.05).

Results:

When comparing implant designs, the all-on-four group (milling [P = 0.002] and SLM [P = 0.001]) exhibited lower marginal gap values than the all-on-six group. No statistically significant difference was observed between the EBM frameworks in both designs. In the all-on-four group, milling resulted in lower marginal gap values than SLM (P = 0.021) and EBM (P = 0.001), while no statistically significant difference was found between the SLM and EBM groups (P = 0.163). For the all-on-six framework design, the milling (P = 0.008) and EBM (P < .001) groups exhibited lower marginal gap values than the SLM group. No statistically significant difference was detected between the milling and EBM groups (P = 0.160).

Conclusion:

Milled frameworks should be the preferred choice for rehabilitations using the all-on-four implant design. For the all-on-six design, both milled and EBM frameworks may be indicated. The marginal gap values observed for all FAFs designs and manufacturing technologies can be considered clinically acceptable.

INTRODUCTION

Implant therapy is widely used to replace missing teeth, demonstrating high success rates.[1] However, the rehabilitation of fully edentulous patients can be influenced by anatomical limitations such as maxillary sinus pneumatization and posterior bone resorption.[2,3] The all-on-four concept was introduced to eliminate the need for sinus elevation and bone grafting – procedures associated with higher costs, longer treatment times, and increased postoperative morbidity.[2,4] By tilting the distal implants (30°–45°), this approach avoids additional surgical interventions while preserving vital structures such as the maxillary sinus.[5] Alternatively, placing short implants in the posterior region provides another solution, as implants shorter than 10 mm are considered a less invasive and more viable treatment option.[6,7] The all-on-six concept, in turn, offers the advantage of eliminating the need for a prosthetic cantilever, thereby reducing surgical complexity.[8] Literature reports indicate that cantilevered prostheses are associated with stress concentration, which increases the risk of biomechanical failures over time.[9,10] Minimizing the cantilever in regions subjected to high masticatory forces may reduce lateral forces and mechanical stress on the prosthesis, improving long-term stability.[8,11]

Assessing the marginal misfit of prosthetic frameworks is essential for verifying manufacturing accuracy.[12] Ideally, the framework should achieve a passive fit, perfectly aligning with the implants and adapting precisely to the supporting implant frameworks, without inducing any stress.[13] Marginal misfit values ranging from 150 μm[14,15] to 230 μm[16] are generally considered clinically acceptable. A high level of accuracy is essential to minimize gaps and mechanical stresses within the implant-support system.[13] Inadequate fit can lead to biomechanical complications, including bone resorption, screw loosening or fracture, and potential failure of the prosthesis.[17,18] Therefore, ensuring a passive fit should be a primary objective in the fabrication of implant-supported prostheses to enhance their stability and longevity.[12,15,19,20]

To minimize errors associated with conventional casting techniques, computer-aided design (CAD) and computer-aided manufacturing (CAM) technologies have been extensively investigated for their potential to improve the dimensional accuracy of implant-supported frameworks.[15,21,22] CAD/CAM technologies can be categorized based on their fabrication approach.[23] The milling technique, a subtractive method, involves shaping a solid block using diamond rotary instruments controlled by a computer-guided machine.[24,25] Studies indicate that milled frameworks provide superior adaptation compared to conventionally cast frameworks.[22,26] However, milling has drawbacks, including limited reproduction of fine details and material waste.[25,27] As an alternative, additive manufacturing (three-dimensional [3D] printing) builds objects layer by layer using powdered or liquid materials, eliminating material waste.[20,22]

Among additive manufacturing techniques, selective laser melting (SLM) and electron beam melting (EBM) are used for printing prosthetic frameworks.[13] In addition, 3D SLM printing has demonstrated satisfactory biomechanical performance in the fabrication of crowns[23,28] and fixed partial dental prostheses.[22,29] However, both SLM and, particularly, EBM are rarely employed in full-arch framework (FAF) fabrication due to the complex 3D geometry required for such rehabilitations. The authors are unaware of a previous study that compared the marginal misfit of FAFs with all-on-four and all-on-six implant designs manufactured by milling, SLM, and EBM. Therefore, this study aimed to compare the marginal gap of FAFs fabricated using all-on-four and all-on-six implant designs. In addition, it evaluated the impact of different CAD/CAM technologies (milling, SLM, and EBM) on marginal fit. The null hypotheses tested were that (1) implant design (all-on-four or all-on-six) and (2) manufacturing technology (milling, SLM, or EBM) would have no effect on the marginal gap of the FAFs.

MATERIALS AND METHODS

A simplified tomography scan of a fully edentulous maxilla from the Renato Archer Information Technology Center (Campinas, São Paulo, Brazil) database was used to create two prototyped master models [Figure 1a and b]. The implant and abutment arrangement (Conexão Sistemas de Prótese Ltd., São Paulo, Brazil) in the master models, along with their specifications, are detailed in Table 1.

Figure 1.

Master model representing a prototyped replica of an edentulous maxilla, illustrating: (a) The all-on-four concept; (b) the all-on-six concept

Table 1.

Positioning and characteristics of the implants in the master models

Implant design	Implant positioning	Implant characteristics	Abutment characteristics
All-on-four	Lateral incisors	(2) Standard EH 4.1 mm × 11.5 mm	Microunit 4.1 mm × 4.0 mm
	Second premolar	(2) Long EH 4.1 mm × 13 mm (tilted 30°)
All-on-six	Lateral incisors	(2) Standard EH 4.1 mm × 11.5 mm	Microunit 4.1 mm × 4.0 mm
	Second premolar	(2) Standard EH 4.1 mm × 11.5 mm
	Second molar	(2) Short EH 5.0 mm × 7 mm	Microunit 4.1 mm × 5.0 mm

EH – External hexagon

For each master model, a FAF was waxed over the abutments. In the all-on-four rehabilitation, the FAF extended from the upper right first molar to the upper left first molar [Figure 2a-c], while in the all-on-six concept, the upper second molars were also included [Figure 2d-f]. A light scanner (Ceramill Map 400⁺; Amann Girrbach, Koblach, Germany) was used to digitize the master models and wax patterns. A .stl file was generated, and the FAFs were designed using specific CAD software (Ceramill Mind; Amann Girrbach). The same CAD file was then sent to the respective machines based on the assigned manufacturing technologies (n = 5).

Figure 2.

Frameworks produced with different three-dimensional technologies and implant number concepts. Note the arch pronounced of the frameworks that simulate the human maxilla. All-on-four frameworks produced by (a) milling; (b) selective laser melting (SLM); (c) electron beam melting (EBM). All-on-six frameworks produced by (d) milling; (e) SLM; (f) EBM

For the milling group [Figure 2a and d], Ti-6Al-4V alloy blocks (89.4% Ti, 6.2% Al, 4% Vn, <0.4% N, <0.4% C, <0.4% Fe, and <0.4% O) (Starbond Ti5; Scheftner, Mainz, Germany) were processed in a five-axis milling machine with an integrated irrigation system (CNC D15W; Yenadent, Istanbul, Turkey). The same Ti-6Al-4V alloy was used in the SLM group [Figure 2b and e], manufactured in an Mlab cusing 200R machine (GE Additive, Cincinnati, Ohio, USA), with a powder composition of 88.47% Ti, 6.5% Al, 4.5% Vn, 0.25% Fe, 0.08% C, 0.13% O, 0.05% N, and 0.012% H. In the EBM group [Figure 2c and f], Ti-6Al-4V alloy was used in the Q10 machine (Arcam GE Additive), with a powder composition of 89.7% Ti, 6% Al, 4% Vn, 0.1% Fe, 0.03% C, 0.15% O, 0.01% N, and 0.003% H. The complete parameters for both the milling and additive manufacturing techniques are presented in Table 2.

Table 2.

Parameters of milling and additive manufacturing machines

Parameters	Milling	Parameters	SLM	EBM
Smallest controllable increment	0.1 μm	Particle size	15–45 μm	45–100 μm
Travel limit A axis	± 28°	Power layer thickness	25 μm	50 μm
X, Y, and Z axis motor powers	400 W	Fiber laser	200 W	3000 W
Accuracy	< 10μm	Focus diameter	75 μm	3.073 mA
Repeatability	< 2μm	Working atmosphere	Argon	High vacuum
Spindle speed range	60,000 rpm	Scan speed	1250 mm/s	4530 mm/s
Maximum power consumption	2.7 kW	Power supply	1.5 kW	7 kW

SLM: Selective laser melting, EBM: Electron beam melting

For gap measurement, the FAFs were passively positioned on the master models and stabilized using a drop of low-shrinkage acrylic resin (GC Pattern Resin; GC America Inc.) applied to the mesial region of each mini-pillar/framework interface, ensuring passive adaptation without external interference. The marginal gap analysis was conducted by a previously calibrated examiner (LDRS) (intraclass correlation coefficient: 0.996; P < 0.001) using a microscope with 1 μm accuracy at ×120 magnification (UHL VMM 100 BT; Walter Uhl, Asslar, Germany) and an analyzing unit (QC 220 HH Quadra-Chek 200; Metronics Inc., Cincinnati, Ohio, USA). Three measurements were taken at the FAF/abutment interface from buccal and lingual views at opposite positions, and an average value was recorded for each FAF. An octagonal support was prototyped to standardize the positioning of the master model and framework during the marginal gap assessment [Figure 3]. In total, 3 steps were followed to fulfill the study design and methodology [Figure 4].

Figure 3.

Prototyped octagonal support stabilizing both the master model and the framework during measurement acquisition. The dotted line indicates possible positions for standard readings. The blue overlay highlights the incidence of light direction during image obtention. All-on-four selective laser melting framework under the microscope

Figure 4.

Schematic representation of the study flowchart. In the first step (1), the master model is shown, where the all-on-four and all-on-six concepts were applied to a simplified maxilla model. In the second step (2), the frameworks (all-on-four and all-on-six) were waxed and subsequently scanned to obtain the.stl file. In the third step (3), the samples were fabricated using milling or three-dimensional printing techniques (selective laser melting and electron beam melting) and then evaluated for the marginal misfit of the prosthetic frameworks. SLM: Selective laser melting and EBM: Electron beam melting

Statistical analysis was performed using the Kolmogorov–Smirnov test to verify data normality. A two-way ANOVA was applied to assess the influence of implant design and manufacturing technology on marginal gap values (α = 0.05). All statistical analyses were conducted using SPSS software (IBM SPSS Statistics v20.0; IBM Corp., New York, USA). The required sample size was calculated using G*Power software (Heinrich Heine University, Düsseldorf, Germany), with a significance level of 5% and a test power of 95%. A minimum of four samples was determined to be necessary for marginal gap analysis.

RESULTS

The implant design (P = 0.004), manufacturing technology (P = 0.004), and the interaction between design and technology (P = 0.001) significantly influenced the marginal gap values [Figure 4]. The milled (P = 0.002) and SLM (P = 0.001) FAFs in the all-on-four configuration exhibited lower marginal gap values compared to the all-on-six design, whereas no statistical difference was observed between the EBM FAFs. Within the all-on-four groups, the milled FAFs demonstrated significantly lower marginal gap values than those produced by SLM (P = 0.021) and EBM (P = 0.001), while no statistical difference was found between the SLM and EBM groups (P = 0.163). In the all-on-six FAF design, the milled (P = 0.008) and EBM (P < 0.001) groups exhibited lower marginal gap values compared to the SLM group, with no significant differences between the milled and EBM groups (P = 0.160) [Figure 5].

Figure 5.

Box plots of marginal gap values (μm) for milling, selective laser melting, and electron beam melting on all-on-four and all-on-six implant designs. Bars indicating statistical difference between groups (P < 0.05, Bonferroni test). SLM: Selective laser melting and EBM: Electron beam melting

DISCUSSION

The influence of implant design (all-on-four vs. all-on-six) and manufacturing techniques (3D printing vs. milling) on the marginal gap of FAFs was evaluated. The first null hypothesis was rejected, as the FAF design significantly affected the marginal gap values. Less extensive FAFs are less susceptible to 3D distortions, which may explain the favorable outcomes observed in the all-on-four group.[17] Previous biomechanical studies[19,30] used laboratory master models without simulating human anatomy, whereas this study incorporated a more pronounced arch in both designs to better replicate clinical conditions. In addition, the greater length of all-on-six FAFs makes them more prone to distortions, potentially compromising marginal adaptation.[14,16]

Two finite element analysis studies[8,10] previously compared the biomechanical behavior of all-on-four and all-on-six FAFs, reporting a more favorable stress distribution in the all-on-six concept. Bhering et al.[8] compared treatment concepts using either four or six implants and applied a unilateral 150 N oblique force to the posterior teeth of the FAF prostheses made from zirconia, cobalt–chromium, and titanium. Their findings indicated that all materials remained within the biological tolerance limits for bone, implants, screws, abutments, and displacement magnitude. From this perspective, since all structures in this study exhibited marginal precision below the critical threshold of 150 μm[14,15] to 230 μm,[16] with the highest recorded value being approximately 100 μm, the assessed FAFs are unlikely to compromise either the implant-related ductile components or the nonductile peri-implant biological structures.[9] Further in silico research is still needed to validate the findings on the passive marginal fit of FAF prostheses supported by four or six implants. From this perspective, Topcu Ersöz and Mumcu[10] analyzed the same materials as Bhering et al.[8] – zirconia, cobalt–chromium, and titanium – while also incorporating the high-performance polymer polyetheretherketone. In this study, titanium alloy was chosen due to its advantages, such as being lighter than zirconia and cobalt–chromium and being compatible with three different manufacturing technologies (milling, SLM, and EBM). In addition, the findings identified the titanium as viable clinical option.[10] While finite element analysis allows for 3D visualization of stress distribution, potentially indicating distortions in implant-supported systems, it is limited to validating the specific models included in the study. As a result, the complex clinical scenario involving implants and FAFs may still lack comprehensive evidence.[21,31]

Previous studies[15,32] compared the marginal gap of milled all-on-four FAFs in zirconia and titanium, demonstrating clinically acceptable values (10–27 μm) for both materials using industrial computed tomography. Katsoulis et al.[19] employed the one-screw test to measure the marginal gap of FAFs supported by six implants, comparing cast cobalt–chromium FAFs to milled zirconia and titanium FAFs. Their findings showed that zirconia and titanium exhibited similar marginal gap values, while cobalt–chromium demonstrated clinically unacceptable results (181–301 μm). The discrepancy between their findings and ours may stem from differences in methodology. The one-screw test protocol[19,22,33] may overestimate marginal gap values, as the unscrewed extremity creates a lever arm, leading to a greater misfit compared to a passively seated FAF, as used in this study.

CAD/CAM technology is renowned for its high precision and consistency, as it minimizes human interference and the potential for errors.[24,25,27] In this study, all-on-four milled FAFs exhibited lower marginal gap values than those fabricated using SLM and EBM. Consequently, the second null hypothesis was rejected, as manufacturing technology significantly influenced marginal gap values. This difference may be attributed to the sequential layering process and the high melting temperatures involved in SLM and EBM fabrication, which impact dimensional precision.[13,25,27] In contrast, the subtractive method currently produces more homogeneous structures, making it a more predictable approach for fabricating prosthetic FAFs.[21,23]

Contrary to our findings, a previous study[22] reported lower marginal gap values for three-unit fixed partial dentures produced through SLM compared to those manufactured by milling and conventional casting. In addition, the same study noted lower standard deviations in SLM-fabricated frameworks. This discrepancy may be explained by differences in framework length, as smaller objects are less susceptible to manufacturing distortions than FAFs.[25] Moreover, FAFs present inherent complexity, with distortions occurring along multiple axes, potentially favoring or hindering final dimensional precision. Despite these challenges, it is important to highlight that the marginal gap values observed in this study were clinically acceptable across all FAF designs and manufacturing techniques.[30,32]

3D printing is an ongoing technology for designing and producing dental prostheses.[13,34] As a novel manufacturing approach, some inaccuracies may still occur. However, it offers advantages such as sustainability, as the powder can be recycled,[29] efficient production of large structures in short-time frames, and the ability to replicate delicate details that milling burs cannot achieve.[24] Nonetheless, standardized protocols for SLM and, particularly, EBM fabrication have yet to be fully established. Factors such as layer thickness, powder particle size, laser speed, and intensity can be adjusted by manufacturers to optimize accuracy and mechanical properties.[25] In a clinical scenario, the decision about which concept to use should be based on anatomical, biomechanical, and financial aspects. The literature shows comparable clinical outcomes between all-on-four and all-on-six designs.[35,36] Despite this, the all-on-four design tends to be more cost-effective, less time-consuming, and less invasive.[35] Regarding patient-reported outcome measures, satisfaction have been reported to vary between the two treatment concepts.[36] The prostheses with four implants were preferred esthetically, while the six-implant option was favored for its superior functional performance, particularly in speaking.[36] Both treatment approaches proved clinically effective.[36] However, the all-on-four is associated with a higher incidence of technical complications and a reduced overall treatment cost.[36] Indeed, the choice for rehabilitation with more implants is a goal for minimizing complications without compromising clinical outcomes.[35]

CONCLUSION

Within the limitations of this in vitro study, despite the complex geometries of FAFs, the marginal gap observed in both implant designs (all-on-four and all-on-six) and across the three manufacturing technologies (milling, SLM, and EBM) remained clinically acceptable (150 μm). Moreover, when selecting the most suitable implant design for each clinical case, considering cost-effectiveness, milled FAFs should be preferred for the all-on-four configuration. In contrast, for rehabilitations supported by six implants, both milled and EBM FAFs are viable options. Therefore, 3D printing emerges as a promising technology for the fabrication of FAFs.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

This work was supported by the Sao Paulo Research Foundation (FAPESP, grant numbers #2017/15297 0, #2018/10378 4, and #2017/16303 3); Coordination for the Improvement of Higher Education Personnel (CAPES) Finance Code 001; Brazilian National Council for Scientific and Technological Development (CNPq, grant numbers #306373/2015-7 and #170040/2018-6); State University of Campinas (UNICAMP) (postdoctoral schoolarship to G.A.B.).