Comparison of the biomechanical properties of CAD/ CAM individualized lingual retainers: An in vitro study

Abstract

Objective

This study examined the biomechanical properties of computer-aided design/computer-aided manufacturing (CAD/CAM) retainers made of titanium and nickel-titanium (NiTi).

Methods

Four retainer types were tested (1) twisted steel, (2) NiTi, (3) cut titanium, and (4) printed titanium. Longitudinal wires were manufactured (2–4) and subjected to a three-point bending test (3PBT). In a biomechanical bending test (BMT), retainers were designed from patient scans and applied to models in an Orthodontic Measurement and Simulation System, with the maxillary central incisor and mandibular central incisor loaded with deflections of ±0.2 mm and rotation of ±1°.

Results

The 3PBT showed that NiTi wires exhibited a characteristic hysteresis loop, whereas the cut and printed titanium wires had different flow and yield strengths. The BMT forces during intrusive and extrusive movements were measured, and the findings revealed that in the maxilla, all CAD/CAM materials showed significant differences from the twisted steel, with NiTi differing from the cut and printed titanium. In the mandible, the forces were generally higher, with significant differences observed at 0.1-mm extrusion and 0.2-mm deflection. In addition, significant differences were observed in the mesial and distal deflections, particularly in the mandible, where the cut titanium generated the lowest forces. Twisted steel exhibited the lowest torque, whereas cut titanium exhibited the highest torque.

Conclusions

CAD/CAM retainers differ significantly from conventional twisted steel retainers, with notable differences between NiTi and titanium retainers, especially in the maxilla. However, these differences may be negligible in the mandible due to the shorter interdental distance.

INTRODUCTION

Retention is key to stabilizing teeth after active orthodontic treatment and preventing relapse due to external forces.¹ After treatment, the results can be maintained with removable or fixed appliances, or a combination of both.² However, many studies have indicated that fixed lingual retainers are superior to removable appliances in preventing relapse.^3,4 Fixed lingual retainers are state-of-the-art devices for tooth stabilization after treatment. The most common material for adhesively bonded lingual retention is a twisted steel wire,⁵ which, in the 0.0155 inch dimension, is considered to have good biomechanical properties.⁶ However, failures and adverse effects are significant risks associated with the use of fixed retainers. Adverse effects such as bonding loss or wire fracture occur at varying frequencies, but their possibility should not be underestimated by the clinician. In particular, the attachment of conventional retainers made of twisted steel wires can be technically demanding and requires skill and experience. The adverse effects of misaligned or incorrectly manufactured retainers include undesired tooth movements, described as the x-effect and twist-effect.^7,8 Therefore, high manufacturing precision and avoidance of small-diameter wires are recommended.⁹ Lingual retainers must resist external forces without restricting tooth mobility, while ensuring fit accuracy, corrosion resistance, and good adhesive adaptation. Computer-aided design/computer-aided manufacturing (CAD/CAM)-fabricated lingual retainers offer an interesting alternative to conventional retainers because of their precise fit, digital workflow, and material variety.¹⁰ Depending on the material, CAD/CAM retainers are cut, robotically bent, milled, or printed either in the office or externally.¹¹ In this regard, CAD/CAM also allows the use of novel materials such as nickel-titanium (NiTi) alloys for retention, with studies suggesting improved retention properties.^12,13 However, other studies have reported higher loss rates for CAD/CAM retainers, especially in the maxilla,¹⁴ with unclear outcomes regarding the influence of the material’s biomechanical properties. Overall, data pertaining to the biomechanical characteristics of these emerging retainers are limited. Therefore, this study aimed to examine the biomechanical properties of CAD/CAM retainers composed of titanium and NiTi in comparison with those of conventional twisted steel retainers.

MATERIALS AND METHODS

Two different tests were performed to compare the retainer materials: a standardized three-point bending test (3PBT) according to ISO standards and a biomechanical bending test (BMT).

All wires used in the 3PBT were 40 mm long. Ten samples of each material were obtained from the same production batch and manufacturer to achieve comparable cross-sectional geometry. The tested dimensions were as follows: Twisted steel retainers (1) were not used for the 3PBT. (2) 0.4 × 0.5 mm for NiTi (CA Digital, Hilden, Germany) (3) 0.6 × 0.5 mm for cut titanium (Teeth & Tech, Wiesbaden, Germany) and (4) 1.0 × 0.5 mm for printed titanium (Medical Additive Manufacturing Laboratory, FH Nordwestschweiz, Muttenz, Switzerland) (Table 1).

Table 1.

Specifications and sample sizes of retainer wires used in the three-point bending test

Material	Dimension	Scaled dimension	Manufacturer	Sample size (n)
NiTi, cut	0.4 × 0.5 × 40.0 mm	0.6 × 0.5 × 40.0 mm	CA Digital, Hilden, Germany	10
Titanium, cut	0.6 × 0.5 × 40.0 mm		Teeth & Tech, Wiesbaden, Germany	10
Titanium, printed	1.0 × 0.5 × 40.0 mm	0.6 × 0.5 × 40.0 mm	Medical Additive Manufacturing Laboratory, FH Nordwestschweiz, Muttenz, Switzerland	10

Scaled dimension indicates the reference cross-sectional size used for normalization (0.6 × 0.5 mm). If the specimen already matches the reference dimension (eg, cut titanium 0.6 × 0.5 mm), no scaling is required and the scaled dimension is left blank.

NiTi, nickel-titanium.

For biomechanical testing using the Orthodontic Measurement and Simulation System (OMSS), eight to ten retainers per group were digitally designed and bonded to a standardized patient model using an intraoral scan. Each material was tested on both the maxillary and mandibular arches, with 8–10 samples per arch (n = 8 for twisted steel, n = 10 for NiTi, n = 10 for cut titanium, n = 10 for printed titanium). Four types of retainers were compared: (1) twisted steel (round, 0.44 mm; Twistflex, Dentaurum, Ispringen, Germany), (2) NiTi (rectangular, 0.35 × 0.35 mm; Memotain, CA Digital), (3) cut titanium (rectangular, 0.60 × 0.50 mm; Titan Retainer, Teeth & Tech), and (4) printed titanium (rectangular, 1.00 × 0.50 mm; FH Nordwestschweiz) (Table 2). Both titanium retainers were made from commercially pure titanium grade 2 (TiCP Grade 2).

Table 2.

Overview of retainer materials used in the biomechanical bending test

Material	Manufacturing process	Dimension	Manufacturer	Sample size (n)
Twisted steel	Conventional	Round, 0.44 mm	Dentaurum, Ispringen, Germany	Maxillary arch: 8 Mandibular arch: 8
NiTi, cut	Cut and laser polished	Rectangular, 0.35 × 0.35 mm	CA Digital, Hilden, Germany	Maxillary arch: 10 Mandibular arch: 10
Titanium, cut	Cut	Rectangular, 0.60 × 0.50 mm	Teeth & Tech, Wiesbaden, Germany	Maxillary arch: 10 Mandibular arch: 10
Titanium, printed	Printed	Rectangular, 1.00 × 0.50 mm	Medical Additive Manufacturing Laboratory, FH Nordwestschweiz, Muttenz, Switzerland	Maxillary arch: 10 Mandibular arch: 10

NiTi, nickel-titanium.

In the mandibular specimens, isolated specimen loss during mechanical testing resulted in reduced sample sizes in two test conditions: for the twisted steel retainer during mesial/distal displacement (n = 7) and for the cut titanium retainer during lingual/labial displacement (n = 3). All available specimens were included in the statistical analysis.

CAD/CAM retainers were designed digitally: the retainers in group II were designed and manufactured by CA Digital; retainers in group III were designed and manufactured by Teeth & Tech using OnyxCeph^TM software (Image Instruments, Chemnitz, Germany); and retainers in group IV were designed and manufactured by the Medical Additive Manufacturing Laboratory (FH Nordwestschweiz).

Three-point bending test

The 3PBT was conducted according to DIN EN ISO 15841 at a standardized temperature of 37°C. A sample wire was placed on two support points 10 mm apart, with a punch applying displacement to the wire center, deflecting it up to 3.1 mm (3.3 mm for NiTi wires) and back (Figure 1A). A deflection of 3.3 mm was selected for NiTi to ensure that the unloading measurements were fully captured within the superelastic plateau. The 3PBT was performed for material groups II–IV (Table 1) on ten wires each. The wires were oriented edgewise, with the smaller dimensions being the height of the wire in the direction of displacement. To confirm the reproducibility of the superelastic curves according to the ISO standards, the NiTi wires were tested twice, resulting in a total of 40 tests. The forces on the loading branch of the force/deflection curve were recorded at deflections of 3.0, 2.0, 1.0, and 0.5 mm. Owing to dimensional differences from manufacturing, test results were converted to a uniform dimension (0.6 mm × 0.5 mm) using the following formula:

Figure 1. Three-point bending test (3PBT) and force-deflection characteristics of different retainer materials.

A, 3PBT setup performed according to ISO 15841 using the Orthodontic Measurement and Simulation System (top view). A 40-mm sample wire was placed upon two supports separated by a distance of 10 mm. A central punch applied force to deflect the wire, and the resulting forces were recorded at predefined deflection points. B–D, Representative force–deflection curves (n = 1) obtained from the 3PBT for different rectangular wires (0.6 mm × 0.5 mm): B, NiTi wire (maximum deflection 3.3 mm); C, Cut titanium wire (maximum deflection 3.1 mm); D, Printed titanium wire (maximum deflection 3.1 mm). E, Force/deflection diagram for 0.1-, 0.2-, 0.3-, and 3.0-mm deflection. F, Graph showing the force/deflection rate for all three retainer materials.

NiTi, nickel-titanium; n.s., not significant.

***P < 0.001, ****P < 0.0001.

$$f_{\text{normalized}}=F_{\text{measured}}x\frac{w_1{h}_1^3}{w_2{h}_2^3}$$

Orthodontic Measurement and Simulation System

The OMSS consisted of two motorized positioning tables equipped with two 6-axis force/moment transducers, which could be operated manually or by computer-controlled three-dimensional movements. Translations and rotations in all three spatial directions were performed with accuracies of 0.001 mm and 0.01°, respectively. Two force/torque sensors mounted on the positioning tables registered the force and torque values.^15,16 The OMSS was connected to a personal computer with a specialized control program that generated predefined movements and simultaneously recorded multidimensional force measurements (N, N·mm). The sensitivity was 0.02 N for forces and 0.5 N·mm for torques. The measuring range was ±15.00 N for forces and ±450 N·mm for torques. For experiments involving temperature-dependent NiTi wires, a constant temperature of 37°C ± 1°C was maintained in the system’s temperature test chamber.

Biomechanical bending tests

Figure 2 presents the experimental setup for the BMT. Retainers composed of four materials were fabricated for the maxilla and mandible: twisted steel, NiTi, cut titanium, and printed titanium (Figure 2A). A universal patient model of the maxillary and mandibular arch was prited using liquid resin (Form3; Formlabs GmbH, Berlin, Germany). The maxillary and mandibular right central incisors were isolated. Each retainer was adhesively bonded to the remaining anterior teeth using a flowable composite (Venus Diamond Flow, Kulzer, Hanau, Germany), without etching or bonding, owing to the direct bond between the acrylic resin and composite. The model was mounted on the positioning table of the OMSS, and the isolated tooth was connected to a force/torque sensor and passively bonded to the retainer (Figure 2B). All the deflections were applied in a standardized sequence. The initial tooth position was carefully verified and manually corrected before each measurement to ensure consistent starting conditions. The OMSS enabled precise computer-controlled deflection and simultaneous measurement of the resulting forces. Deflections included translation along the x-axis (intrusion and extrusion, ±0.2 mm), y-axis (lingual and labial shift, ±0.2 mm), and z-axis (mesial/distal shift, ±0.2 mm) and rotation around the x-axis (±1°) of the isolated incisor (Figure 2C).

Figure 2. Representation of the experimental setup and the different retainer materials.

A, Photographs of the retainers used in this study. From left to right: twisted steel, nickel-titanium (NiTi), cut titanium, and printed titanium. B, Biomechanical bending test in the Orthodontic Measurement and Simulation System: top view of the printed model with a bonded retainer and isolated tooth (lower central incisor) before bonding to the retainer. C, Coordinate system displaying directions of deflection applied to an isolated tooth in the biomechanical bending test.

Data evaluation and statistical analysis

All data were analyzed using Excel 16.44 (Microsoft Excel, Redmond, WA, USA), SPSS 24 (IBM Corp., Armonk, NY, USA), and GraphPad Prism 10 (GraphPad Prism Inc., La Jolla, CA, USA). The data were transferred from an OMSS Linux computer to Excel. Normality was assessed using the Kolmogorov–Smirnov test in SPSS, which showed a normal distribution for all measurements except for rotation around the x-axis (R(x)). Three-point bending test data were analyzed using univariate analysis of variance (ANOVA) with post hoc Scheffé tests for mean comparisons. The Kruskal–Wallis test was used for R(x), followed by post hoc Scheffé tests. The results of the BMT were analyzed using a two-way ANOVA, followed by Tukey’s multiple comparison analysis. Given the reduced sample size in one mandibular lingual/labial subgroup (n = 3), variance homogeneity was assessed (Brown–Forsythe, P > 0.05). Welch-corrected and sensitivity analyses confirmed robustness of the results. Statistical significance was set at P < 0.05. The results were visualized using the GraphPad Prism 10.

RESULTS

Three-point bending test

Figure 1B–D shows the force/deflection diagrams for representative samples (n = 1) of groups II–IV. The NiTi wire (Figure 1B) exhibited a characteristic hysteresis loop with higher forces during the loading phase than during the unloading phase (Figure 1B). Each wire was subjected to two loading/unloading cycles, resulting in two lines per measurement point. Figure 1B shows the characteristic force-displacement diagram of NiTi. The loading plateau started at 7.99 ± 0.16 N and 0.83 ± 0.05 mm deflection. The maximum force was 9.09 ± 0.11 N and corresponded approximately to the force at the middle of the loading plateau (9.17 ± 0.28 N). After the maximum deflection of 3.3 mm, the loading plateau changed to the unloading plateau, with a mid-force of 5.67 ± 0.12 N. Slight plastic deformations occurred in all NiTi samples.

To further characterize the biomechanical properties of the cut (3) and printed (4) titanium test wires, we determined their flow limit and yield strength. Overall, the printed wires (4) required less force: The flow limit, the force at which plastic deformation begins, was 7.77 ± 0.49 N for cut titanium and 5.03 ± 0.20 N for printed titanium. The yield strength, which describes the onset of permanent deformation beyond which the retainers could no longer return to their original shape, of group III was 4.18 ± 0.35 N and that of group IV at 3.47 ± 0.20 N. Because NiTi exhibits a pseudoelastic behavior, these three materials can only be compared to a limited extent. Comparing all the wires tested, the NiTi wires (2) showed a force of 2.35 ± 0.20 N at 0.2-mm deflection; the cut titanium wires (3) showed a force of 2.60 ± 0.27 N; and the printed titanium wires (4) showed a force of 1.02 ± 0.10 N. These represent statistically significant differences among printed titanium, NiTi, and cut titanium, which were not observed at a 0.1-mm deflection (Figure 1E). With increasing deflection, NiTi and cut titanium remained at markedly higher force levels than printed titanium Supplementary Figure 1). This trend was evident throughout the loading range (0.5–3.0 mm deflection). At 0.5-mm deflection, mean forces were 4.24 ± 1.76 N for NiTi, 5.36 ± 0.63 N for cut titanium, and 2.55 ± 0.26 N for printed titanium. With increasing deflection, all materials showed rising force values, reaching 9.05 ± 0.20 N for NiTi, 7.35 ± 0.51 N for cut titanium, and 5.71 ± 0.55 N for printed titanium at 3.0 mm. Overall, NiTi and cut titanium demonstrated markedly higher resistance to bending than printed titanium, with statistically significant differences emerging from 0.2 mm deflection onward (P < 0.0001).

The gradient indicates the stiffness of the material, with a higher gradient indicating stronger resistance to bending. The gradient values were 11.31 ± 0.20 N/mm for group II, 15.54 ± 1.51 N/mm for group III, and 5.26 ± 0.39 N/mm for group IV (Figure 1F).

Biomechanical bending test

Figure 3A and 3B show an overlay of intrusive and extrusive force measurements on the central maxillary (Figure 3A) and mandibular (Figure 3B) incisors for all four retainer materials. The corresponding deflection diagrams (Figure 3C and 3D) show forces within a ±0.2-mm range, with positive values indicating extrusion and negative values indicating intrusion. In the maxilla (Figure 3C), all CAD/CAM materials significantly differed from twisted steel at 0.2-mm intrusion. At 0.2-mm extrusion, twisted steel differed from all CAD/CAM materials, and NiTi differed from both cut and printed titanium (Figure 3C). No significant difference was found between cut and printed titanium. In the mandible, no significant differences were observed at up to 0.2-mm intrusion (Figure 3D). At 0.1-mm extrusion, printed titanium differed significantly from twisted steel and NiTi, and at 0.2-mm extrusion, NiTi differed from both cut and printed titanium (Figure 3D). Notably, forces in the mandible were greater than those in the maxilla (Figure 3).

Figure 3. Force-deflection characteristics of central maxillary and mandibular incisor under intrusive and extrusive loading.

A, Force/deflection diagram of intrusive and extrusive forces generated on the maxillary central incisor with a deflection range of ±0.20 mm in the biomechanical bending test. B, Force/deflection diagram of intrusive and extrusive forces generated on the mandibular central incisor with a deflection range of ±0.20 mm in the biomechanical bending test. C, Boxplots of the resulting forces during the biomechanical bending test depending on intrusion and extrusion of 0.10 mm and 0.20 mm of the maxillary central incisor (n = 8–10). D, Boxplots of the resulting forces during the biomechanical bending test depending on intrusion and extrusion of 0.10 mm and 0.20 mm of the mandibular central incisor (n = 8–10).

NiTi, nickel-titanium.

*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4A and 4B show mesial and distal force measurements on the central incisor for all four retainer materials, with forces displayed across a ±0.2-mm deflection range. Positive values represent mesial movements, and negative values represent distal movements. In the maxilla, no significant differences were found among the materials. However, in the mandible, significant differences were observed between twisted steel and cut titanium and between NiTi and cut titanium at 0.2-mm mesial deflection. At 0.2-mm distal shift, printed titanium significantly differed from twisted steel and cut titanium. Cut titanium generated the lowest forces for both mesial and distal deflections (Figure 4D). As with previous findings, forces in the mandible were consistently higher than those in the maxilla (Figure 4), indicating biomechanical variations between the both arches.

Figure 4. Force-deflection characteristics of central maxillary and mandibular incisor under mesial and distal loading.

A, Force/deflection diagram of displacement to the mesial and distal directions generated on the maxillary central incisor at a deflection range of ±0.20 mm in the biomechanical bending test. B, Force/deflection diagram of displacement to the mesial and distal directions generated on the mandibular central incisor at a deflection range of ±0.20 mm in the biomechanical bending test. C, Boxplots of the resulting forces during the biomechanical bending test depending on distal (–0.10, –0.20 mm) and mesial (+0.10, +0.20 mm) displacement of the maxillary central incisor (n = 8–10). D, Boxplots of the resulting forces during the biomechanical bending test depending on distal (–0.10, –0.20 mm) and mesial (+0.10, +0.20 mm) displacement of the mandibular central incisor (n = 7–10).

*P < 0.05, ***P < 0.001, ****P < 0.0001.

Figure 5A and 5B show lingual and labial force measurements on the central incisor for all four retainer materials, with forces displayed across a ±0.2-mm deflection range. Positive values correspond to labial movements, and negative values correspond to lingual movements. In the maxilla, no significant differences were found in the 0.1-mm labial deflection. At a labial deflection of 0.2 mm, all materials showed significant differences from twisted steel and NiTi (Figure 5C). At a lingual deflection of 0.1 mm, cut titanium differed significantly from NiTi and printed titanium, while at a lingual deflection of 0.2 mm, cut titanium significantly differed from all other materials (Figure 5C). In the mandible, printed titanium differed significantly from NiTi at a 0.2-mm lingual deflection, and NiTi differed significantly from twisted steel (Figure 5D). The forces in the mandible were consistently higher than those in the maxilla (Figure 5), reinforcing the trend of greater forces in the mandible jaw.

Figure 5. Force-deflection characteristics of central maxillary and mandibular incisor under lingual and labial loading.

A, Force/deflection diagram of displacement to lingual and labial directions generated on the maxillary central incisor at a deflection range of ±0.20 mm in the biomechanical bending test. B, Force/deflection diagram of lingual and labial displacements generated on the mandibular central incisor at a deflection range of ±0.20 mm in the biomechanical bending test. C, Boxplots of resulting forces during the biomechanical bending test depending on lingual (–0.10, –0.20) and labial (+0.10, +0.20) displacements of 0.10 mm and 0.20 mm of the maxillary central incisor (n = 8–10). D, Boxplots of resulting forces during the biomechanical bending test depending on the lingual (–0.10, –0.20) and labial (+0.10, +0.20) displacements of 0.10 mm and 0.20 mm of the mandibular central incisor (n = 3–10).

*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6A and 6C show rotational measurements on the central incisor for all four retainer materials, with torque (in N·mm) generated across a ±1.0° rotation range. Positive values represent the initial rotational measurements. In the maxilla, twisted steel exhibited a significantly lower torque than the other materials and maintained the lowest torque in the reverse movement (Figure 6C). In the mandible, twisted steel showed the lowest torque, differing significantly from both NiTi and cut titanium, whereas cut titanium produced the highest torque and differed significantly from both NiTi and printed titanium (Figure 6D). During reverse rotation, all materials exhibited lower torques, with twisted steel differing significantly from the cut and printed titanium (Figure 6D). Notably, the torque values in the maxilla were higher than those in the mandible (Figure 6).

Figure 6. Torque-deflection behavior of central maxillary and mandibular incisor during rotational displacement.

A, Torque/deflection diagram: resulting torque of rotation on the maxillary central incisor with a rotation range of ±1.0° in the biomechanical bending test. B, Torque/deflection diagram: resulting torque of rotation on the mandibular central incisor with a rotation range of ±1.0° in the biomechanical bending test. C, Boxplots of the resulting torques during the biomechanical bending test depending on the rotation of –1° and 1° of the maxillary central incisor (n = 8–10). D, Boxplots of the resulting torques during the biomechanical bending test depending on the rotation of –1° and 1° of the mandibular central incisor (n = 8–10).

*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

DISCUSSION

This study aimed to evaluate and compare the biomechanical properties of CAD/CAM-fabricated retainers made of titanium and NiTi with those of conventional twisted steel retainers. The results from the 3PBT and the BMT showed that the tested CAD/CAM retainers made from titanium and NiTi performed differently from each other and from conventional retainers made of twisted steel. Thus, in addition to differences in the retainer material, those in the manufacturing process may influence the mechanical properties of retainers, with noticeable differences between printed and cut titanium CAD/CAM retainers.

Taken together, the results of the 3PBT confirmed that both the material and manufacturing process influence the biomechanical properties of the retention wire. Differences in forces and biomechanical properties were observed among the three wire products. The NiTi wires exhibited distinguishable loading and unloading phases with a typical hysteresis curve, which returned to the baseline and demonstrated superelastic properties.^17,18 In contrast, the titanium wires exhibited permanent deformation, with the wires not returning to their original state. The yield point for the printed titanium was reached at a lower force than that for the cut titanium.

In the clinical context, this is particularly relevant because a high yield point indicates that a retainer retains its properties even after repeated deformations, implying long-term durability and effectiveness. The lower yield point of the printed titanium may affect its performance over time. In particular, materials that effectively maintain their form are favorable for long-term retention and can prevent frequent adjustments. Although 3PBT offers high reproducibility¹⁸ and provides valuable biomechanical insights, it does not replicate intraoral conditions.¹⁷

The OMSS analysis in this study replicated the biomechanical behavior under clinical conditions.^15,16 In contrast to the results obtained by Möhlhenrich et al.,¹⁷ our study included retainers in both arches, considering the distinct biomechanical properties of the maxillary and mandibular arch. Higher clinical failure rates of maxillary arch retainers have been consistently reported.^19-22

Notably, the study by Jowett et al.¹⁴ first reported high failure rates of CAD/CAM NiTi retainers in the maxillary arch, whereas retainers in the mandibular arch showed no difference in failure rates. This is in contrast to the results of previous studies, which generally highlighted the favorable properties of NiTi retainers for the maxillary arch.²⁰ Several studies have linked the higher loss rates of maxillary-arch retainers to shear forces from bite pressure, arch form, and maxillary canine geometry. An increase in the post-retention overbite may further contribute to canine trauma and material fatigue.^23,24 However, most studies have not assessed the biomechanical properties of these materials.

During our OMSS investigation, the differences between the materials were more pronounced in the maxillary arch, as shown by the intrusion/extrusion and lingual/labial displacements. In the mandible arch, the differences between the retainer groups were smaller, but the overall forces were higher than those in the maxillary arch. This is due to the narrower teeth, which resulted in a smaller distance between the attachments and produced a higher force-displacement rate comparable to the distance between the brackets. Conversely, rotational forces were greater in the maxillary arch, where a longer inter-attachment distance created a longer lever arm, amplifying material differences and potentially explaining the higher failure rates observed in the maxillary retainers.

Although previous studies have demonstrated a higher failure rate for maxillary-arch retainers, they have also found that rebonding did not cause unwanted tooth movement, enabling successful reattachment.¹⁴ A randomized controlled trial^25,26 found no significant difference in the mandibular anterior stability between CAD/CAM NiTi and twisted steel retainers. Both types were equally effective in preserving periodontal health, emphasizing that the CAD/CAM technology can be used without compromising clinical outcomes.^27-29 The clinical influence of titanium retainers on the stability of mandibular incisors and periodontal health was further corroborated by Seki Yurdakul and Meriç.³⁰

However, the optimal material selection must also consider patient comfort, hygiene, and integration into the clinical workflow. Studies have shown that smooth CAD/CAM surfaces accumulate less plaque and are easier to clean than twisted wire surfaces.^27,29 Additionally, the ease of integrating the workflow into daily clinical practice is important. The digital workflow of CAD/CAM retainers integrates well into daily practice, especially because of the highly accurate placement results.³¹ This is particularly relevant for maxillary retention, where precise planning and accurate placement are critical for retainer success.³²

A limiting factor of this study was the absence of physiological conditions. Intraoral factors such as saliva, temperature changes, and oral pH may affect the long-term performance and durability of retainers.^15,16 The periodontal ligament also plays a crucial role in the biomechanical behavior and long-term stability,^24,33,34 especially when attachment loss causes greater tooth movement. Further limitations resulted from the study methodology. Although the retainers were bonded to the models by an experienced practitioner under standardized conditions, minor variations in wire fixation could still have influenced the OMSS results. In addition, anatomical and material anisotropies may have contributed to the asymmetrical force patterns observed during the OMSS. The mechanical behavior of metallic materials is strongly influenced by their geometry and manufacturing processes. Factors such as the helical structure of the twisted steel wires and process-related effects, including the rolling of titanium or layer-by-layer buildup in additive manufacturing, can lead to anisotropic properties.^35-37 For cut titanium or NiTi components, the cutting process itself may alter the local microstructure depending on its thermal and mechanical effects. Comparable direction-dependent effects have also been observed during ultraprecision cutting of single-crystal materials such as silicon.³⁸ These examples illustrate that the anisotropic behavior of a material often results from a combination of geometric design and processing history, which could partially explain the asymmetric force responses observed in the OMSS.

In addition, as part of the 3PBT, all retainer materials were scaled to uniform dimensions to assess their basic biomechanical properties. In OMSS trials focusing on clinical applicability, differences in manufacturing processes, dimensions, and geometries complicate direct material comparisons. The higher forces in cut titanium retainers in comparison with NiTi wires are partly attributable to their larger diameter, indicating dimensional variations from the fabrication limit comparability. While retainer dimensions do not affect retention efficiency,³⁹ they substantially affect mechanical properties.⁴⁰

Another challenge in interpreting our findings is the limited availability of clinical data on the arch-dependent failure rates of different retainer materials, since most existing studies have focused primarily on the mandibular arch, hindering direct comparison with our findings.

To the best of our knowledge, this is the first study to experimentally compare alloys fabricated using two manufacturing processes. Future clinical studies should focus on the long-term stability, biomechanical performance, and comfort of different retainer materials.

CONCLUSIONS

In conclusion, CAD/CAM retainers exhibited significant differences from conventional twisted steel retainers, particularly between the NiTi and titanium versions. These differences were particularly pronounced in the maxilla, where they may contribute to higher failure rates. However, these differences appeared to be minimal in the mandible owing to the smaller interdental distances. In the clinical context, this leads to the conclusion that, in the mandible, the choice of CAD/CAM retainer material appears to have minimal impact, allowing for the flexible use of different materials. In the maxilla, owing to its greater biomechanical variability, more frequent follow-ups are advised, since the choice of an inappropriate material may be associated with higher failure rates.

Nevertheless, further clinical studies are needed to evaluate the efficacy of CAD/CAM retainers with different material properties in the maxilla since existing research remains limited.

SUPPLEMENTARY MATERIAL

Supplementary data is available at https://doi.org/10.4041/kjod25.114