The efficacy of polyether‐ether‐ketone wire as a retainer following orthodontic treatment

Abstract Objectives To investigate the efficacy of polyether‐ether‐ketone (PEEK) wire as a fixed orthodontic retainer, by comparing its performance to other retainer wires and optimizing its adhesion to composite bonding materials. Materials and methods Retainer wires of 15 mm segments were used, PEEK wires were prepared in cylindrical form with 0.8 mm diameter, and had two surface treatments namely air‐abrasion and conditioning with adhesive system. Three different metallic retainer wires were used for comparison and three tests were performed; two tests measured debonding force and associated wire deflection from acrylic blocks and bovine teeth and one test for pull‐out force. To test debonding force, a vertically directed compressive force was applied to the retainer wires bonded to the acrylic blocks and bovine teeth, while for pull‐out test; a vertically directed tensile force detached the retainer wire. Results In both debonding tests, PEEK wires (regardless the surface treatment) had non‐significant difference when compared to each other, or to the other metallic wires, except the dead‐soft coaxial wire group. The dead‐soft coaxial wire group had significant difference when compared to other groups regarding both the force magnitude and maximum deflection, the only exception was the debonding force of the flat braided retainer wires bonded to bovine teeth. In pull‐out test PEEK wires conditioned with adhesive system and the air‐abraded recorded the second and third highest readings respectively. Conclusions Within the limitations of this study, the 0.8 mm round PEEK wires have comparable performance—in terms of debonding and pull out forces—to conventional retainers when bonded with 4 mm composite bonding spots; using air‐abrasion for 10 s at 3.5 MPa provided sufficient adhesion of the composite to the wire, and conditioning with adhesive system may provide no further clinical benefit.


| INTRODUCTION
Retention is an integral step of orthodontic treatment; it involves holding the teeth in their post-treatment position, so that excellent long-term result is obtained. Whenever intra-arch instability is suspected, prolonged retention must be considered (Proffit et al., 2019). The long survival time of fixed retainers is one of the reasons that makes them the first choice for prolonged retention (Jin et al., 2018). Failure of bonded retainers, may occur in one of three sites, the wire composite interface, the adhesive-enamel interface, and the wire itself. Failures at the first site are mostly related to composite (Bearn, 1995); failures at the second site are mostly attributed to moisture contamination or retainer wire movement during bonding (Butler & Dowling, 2005); lastly, failures at the third site are the result of stress-fracture and are mostly associated with small diameter wires (Shaughnessy et al., 2016), and long service time (Lumsden et al., 1999). Failures will lead to changes in post treatment tooth position, that is, failed retention, either toward the pre-treatment position, i.e. relapse, or to a totally new position (Butler & Dowling, 2005). However, failure of retention does occur sometimes without involving retainer failure (Sifakakis et al., 2015). This failure is the result of the development of force components within the wire as a result of improper contouring and adaptation at the time of bonding, or deformation developed due to masticatory forces for example (Shaughnessy et al., 2016;Sifakakis et al., 2011). Retention failure is not limited to a single type of fixed retainer, it can be seen with small diameter dead-soft wires, small diameter flexible spiral wires, large diameter stiffer wires, and fiber-reinforced composite (Sfondrini et al., 2020;Shaughnessy et al., 2016).
PEEK is attracting attention as an alternative to metal alloys in the dental field (Tada et al., 2017). It is a high-temperature semi-crystalline, and thermoplastic polymer, that is known for outstanding mechanical properties (Kurtz, 2019). PEEK shows lower deformations and higher fracture loads when processed by computer-aided design/ computer-aided manufacturing (CAD/CAM) than can be achieved by other processes . Good bonding to adhesives is an important requirement for a retainer wire (Annousaki et al., 2017). PEEK has known issues with bonding due to its inert behavior (Stawarczyk, Thrun, et al., 2015); therefore, it has been suggested to use surface treatment with chemical adhesion, micromechanical retention or a combination of both to increase bonding of PEEK to adhesives (Piwowarczyk et al., 2004). According to Caglar et al. (2018) the use of adhesive systems that contain methylmethacrylate monomer resulted in higher bond strength between PEEK and resin. The highest tensile bond strength to PEEK was achieved when the adhesive system Visio.Link (VL), which contains methylmethacrylate monomer, was combined to air-abrasion .
Since there is no high-quality evidence to support endorsing a specific type of retainer as the best retainer (Al-Moghrabi et al., 2016); this study aimed to investigate the efficacy of PEEK wire as a fixed orthodontic retainer, by comparing its performance to other retainer wires and optimizing its adhesion to composite bonding materials.

| MATERIALS AND METHODS
used in the standard mode with 1,000 mW/cm 2 to investigate whether curing time of VL has an effect on the debonding force. The test procedures were identical to that followed in debonding test from acrylic blocks described later. The ultimate force of failure, which is the highest force scored throughout the test, was recorded for each sample.

| Debonding from acrylic blocks
A specially designed and milled brass mold was used to obtain acrylic blocks. The block consisted of two parallel arms, 7 mm in width for each, separated by a 2 mm space, to replicate the condition intraorally between the two composite bonding spots of the lower incisors. A guide made of 1 mm stainless-steel wire was used to standardize the position of holes in the arms of the block. A large round bur (1.5 mm in diameter) was used to drill two holes of 2 mm depth through the guide. The drilled holes were cleaned off acrylic flakes and became ready for bonding ( Figure 1).
Six groups of 10 lingual retainer wires were prepared (same grouping for all three tests): • Group I (NT): PEEK wires with no surface treatment. Cleaned under running water, rinsed in 40% alcohol to remove debris, and dried.
• Group II (AA): PEEK wires that were air-abraded for 10 s at a 10 mm distance (Supporting Information), with AL 2 O 3 50 μm particles at 0.35 MPa. Cleaned under running water, rinsed in 40% alcohol to remove debris, and dried.
• Group III (AV): PEEK wires that were air-abraded for 10 s at a 10 mm distance, with AL 2 O 3 50 μm particles at 0.35 MPa; cleaned under running water rinsed in 40% alcohol to remove debris, dried, then conditioned with a thin layer of VL and cured for 20 s. The head of the curing unit was held as close as was possible, and curing was done against a mirror, in as close as 1 mm between the wire and the mirror, to provide curing for both sides of the wire at the same time (Supporting Information).
Groups IV, V, and VI were cleaned under running water, rinsed in 40% alcohol to remove debris, and dried.

| Debonding from bovine teeth
For debonding force from bovine teeth, the roots were cleaned with a blade and a curette to remove remnants of the periodontal tissues.
The teeth were then stored in a 1% thymol solution. Since bovine incisors had a more triangular shape than human teeth, it was necessary to cut through one side of the teeth, to have intimate line-contact between them. The trimmed teeth were initially fixed to each other using a rapid setting cyanoacrylate adhesive at the root area ( Figure 4). A custom-made T-shaped tool was used to ensure that the lingual surfaces of the attached teeth were at the same line.
Pairs of teeth were placed in a mold and temporarily fixed in place using soft wax, making sure that the line of teeth contact was perpendicular to the block. The T-shaped tool was attached to a surveyor and used to check that the lingual surface of the pair of teeth was well-aligned and parallel to the side of the block (Supporting Information).
Cold cure acrylic was poured to form the blocks. The lingual surface of each tooth was cleaned with pumice, washed with distilled water, and subsequently dried with air. The lingual enamel surfaces were then etched with 35% orthophosphoric acid solution for 30 s (Ultra-Etch; Ultradent Products Inc., South Jordan, Utah), washed thoroughly with water and finally dried. The primer was applied to the tooth surface without curing, the wire was bonded with the light cure adhesive on the two incisors. It was possible to provide a slight curve to the metallic wire retainers, but this was not possible with the PEEK F I G U R E 2 Acrylic blocks with the bonded PEEK wire. (a) Frontal view. (b) Top view F I G U R E 3 The debonding procedure was continued until the wire ruptured (a) or the composite fractured (b) F I G U R E 4 (a) Bovine teeth were trimmed from the sides (b) to provide line contact between the teeth. (c) Cyanoacrylate was used to fix the teeth in position wires as they were machined in straight pieces. The amount of adhesive was standardized with the use of molds, then cured for 20 s. The test samples were kept in distilled water for 24 h before running the test. The debonding procedure was identical to what has been described previously, and maximum deflection before final failure was also recorded at 30 N ( Figure 5).

| Pull-out test
To perform pull out test, blocks were prepared of acrylic with a hole of 2 mm in width, and 3 mm in depth, drilled at the center of an acrylic block. The hole was cleaned off remnants of acrylic. The prepared wire was placed parallel to the long axis of the block, at the center of the holes, and checked with a surveyor from all sides. The adhesive was cured for 20 s. The test samples were kept in distilled water for around 24 h before running the test ( Figure 6).
The block and wire were aligned so that the pull was along the long axis of the block (Figure 7). The crosshead speed was set to 10 mm/min. The ultimate force of failure was recorded for each sample as the highest force recorded during a pull-out test. The test was continued until either the wire failed by rupturing, or got detached from composite.

| STATISTICAL ANALYSIS
Statistical analyses were performed using the SPSS 26.0 (SPSS Inc., Chicago, IL). The Shapiro-Wilks normality test and Levene's variance homogeneity test were applied to the data. For normally distributed data with equal variances, one-way ANOVA and Tukey HSD post hoc multi-comparison test were used. While those with unequal variances Welch F test and Games-Howell post hoc multi-comparison test were used.

| RESULTS
All data were normally distributed, with unequal variances with the exception of the ultimate force of failure of retainers bonded to bovine teeth. The p-value ≤0.05 was considered significant; however, as a result of multiple testing-measuring force and deflection at the same time-Bonferroni correction was applied with a p-value ≤0.025.

| Scanning electron microscope imaging
Scanning electron microscope images taken before air-abrasion showed a distinctive pattern on the wire surface, which was attributed to the cutting burs of the CNC machine. The surface was relatively uniform with some elevations and depressions. The images F I G U R E 5 The debonding procedure on retainers bonded to bovine teeth F I G U R E 6 The block used in the pull-out test (a), a hole drilled in the center of the block (b), with the retainer wire bonded (c) F I G U R E 7 The test block is attached to the movable upper jaw, while the wire is firmly grabbed by the lower fixed jaw taken after air-abrasion showed total loss of that pattern, with deep grooves and highly irregular surface (Supporting Information).

| Pilot study
Curing VL for 20 s was sufficient to develop a dry surface. No significant difference was found between the three groups; so, it was decided to use the 20 s curing light time for VL in the remaining tests of this study (Table 1 and Supporting Information).

| Debonding from acrylic blocks
There was a statistically significant difference between groups (Table 2), post hoc multi comparisons showed a significant difference between the ultimate force of failure of Re as compared to other retainer wires. PEEK wires (regardless the surface treatment) had non-significant difference when compared to each other, or to the other two metallic wires. The BR had a significant difference when compared to FD. The highest ultimate force of failure of Re was associated with the greatest deformation of the wire before failure (Figures 8-10). When the maximum deflection data was inspected, there was a significant difference between groups, that was between the maximum deflection of Re and all other wires (Table 3).

| Debonding from bovine teeth
The ultimate force of failure differed significantly between the different groups. The Re group had significantly higher ultimate force of failure than all other groups, except BR. There was a non-significant difference among all other groups. All retainer-wire failure was by wire rupture, except BR, where the failure was by composite fracture ( Figure 11). The Re group also had the greatest maximum deflection, which was significantly different when compared to all other groups.

| Pull-out test
There was a statistically significant difference when the pull-out force data was inspected. For the PEEK groups, AV had a statistically significant higher pull-out force when compared to NT and AA. The FD group had a significantly lower ultimate force of failure when compared to all groups. During the test, the AA and AV groups had the wire ruptured during the test; while the wire came out of the block for all other groups (Figure 12).

| DISCUSSION
Most patients would require a fixed lingual retainer to keep their anterior teeth stable after orthodontic treatment (Renkema et al., 2008).
The PEEK is known for its remarkable mechanical properties. It can be modified with additives, like carbon fibers-reinforced, hydroxyapatite, barium sulfate, and titanium dioxide, to change its biomechanical properties and behavior (Kurtz, 2019). PEEK compounds, with and without additives, have the good fatigue resistance properties to withstand cyclic loading. Titanium dioxide increases the stiffness of PEEK, in addition to its original use as a whitening agent (Schwitalla et al., 2017). It was shown that PEEK with 20% TiO 2 filler particles had the highest bond strength when compared to a higher or a lower filler content (Lümkemann et al., 2017).
The mechanical behavior of the wires was investigated by performing three tests. The first test was intended to measure the force required to break the bond between the wire and the composite adhesive through the application of a compressive force. Fixing the wire to the U-shaped acrylic blocks with the composite resin secured to a hole within the block, ensured eliminating bonding to enamel as a variable; also having the plunger passing freely in the slot of the block without any interference, reflected the actual bonding force. The second test was adopted from the works of Cooke and Sherriff (2010) and Baysal et al. (2012), who investigated the combined effect of bonding of the adhesive to the wire and the bond to enamel. The third test involved measuring bonding strength while the retainer wire is subjected to a tensile force, it was adopted from the works of Bearn et al. (1997) and Baysal et al. (2012).
Surface treatment is necessary to increase bonding to PEEK as it is an inert material. Air-abrasion results in roughening of the surface, which will aid in a greater surface area and more interlocking between the PEEK and the adhesive. Moreover, with VL, the presence of pentaerythritol triacrylate modifies the PEEK surface and could result in better bonding with the resin adhesives (Caglar et al., 2018). Since there has been no previous research on bonding to PEEK in the wire form, it was decided to investigate the effects of both surface treatments on debonding force as compared to the non-treated sample.
The performance of PEEK was compared to three retainer wires; a commonly used dead-soft six strand coaxial stainless-steel wire, a braided stainless-steel retainer wire with a flat profile that helps retaining the final torque positions and reduce occlusal interference, and a third single-strand titanium wire, that has, in addition to the features of the braided stainless steel wire, a reduced wear rate and is nickel-free.

| PILOT STUDY
According to the manufacturer's instruction, curing VL requires an ultraviolet band between 370 and 400 nm. Valo LED curing units cover a range of 385-515 nm according to the official website of the product (Ultradent, 2017). There was no data regarding the required time for curing VL using Valo, and whether that time could affect the strength of bonding to PEEK; consequently, a pilot study was performed.
The results of the pilot study showed that there is a nonsignificant difference between the three light-curing times, 20, 40, or 60 s. Hence, it was decided to use the 20 s as the light-curing time of VL for the remaining tests of this study.

| The debonding force of retainer-wires bonded to acrylic blocks
The NT and AA groups had very close results. The results of AV, on the other hand, were greater in the magnitude of the ultimate force of failure, and smaller in the maximum deflection before failure, which could reflect a more rigid behavior; however, these differences were not statistically significant. The pattern of failure for AA and AV groups was due to wire rupture, with minor composite fractures. Both can be attributed to its high ductility.
There is no research with a similar laboratory set up to provide a direct comparison of the results. However, Re wire groups were associated with a higher ultimate force of failure and maximum deflection before failure (Milheiro et al., 2012).
The more deformation before a failure happens, the more the possibility that the teeth are moved to a new position without having the wire ruptured or composite fractured, this is mostly the case with Re group.
7.2 | The debonding force of retainer-wires bonded to bovine teeth Several non-human teeth have been used in research, these include swine, equine, bovine, and shark teeth (Yassen et al., 2011). Bovine lower incisors are much easier to collect and are an inexpensive substitute to human incisors, do not have carious lesions or other defects that may affect test outcomes (Oesterle et al., 1998;Yassen et al., 2011). In their systematic review and meta-analysis,

| The pull-out test
The pull-out test helped to further unveil the differences among the retainer groups. It is known that the greater the surface area the greater the force required to remove (Bearn et al., 1997). The FD group had a significantly lower force of failure as compared to all other retainer groups. This can be mainly attributed to the fact that it is a flat and single-stranded wire; therefore, its surface area is smaller than the closest match, the BR group, which despite having the same dimensions, but being made of three strands of annealed stainlesssteel wires has added to the surface area, and provided more gaps for the resin adhesive to bond. The Re and NT groups were in the lower rank when the pull-out force is concerned, with a non-significant difference between them. Even though the Re group had a smaller diameter, but its multistrand nature would add to the surface area and contribute to the retention of adhesive to the wire. The other PEEK groups namely AA and AV had greater pull-pull out force, this again highlights the importance of the surface area, because air-abrasion resulted in greater surface irregularities for the resin composite to lock-in, and the effect of the adhesive system VL seemed to increase the bonding force. However, it seemed that air-abrasion scratches may result in some weakening of the wire, therefore both groups have ruptured at slightly higher pull-out forces, while the NT group came out intact. Even though the increase in pull out force between the AV and AA groups is statistically significant, it seems of little clinical significance, especially when we consider the failure pattern for both which is wire rupture.

| Limitations of the study
This study has certain limitations such as; lacking a specific protocol to cutting straight pieces of PEEK in a small round cross-section.
The protocol of cutting PEEK for crown and bridge work was modified to get the retainer wire pieces. The manual finishing of the PEEK wires, although it is believed that the effect of this variable is limited by keeping the variation within 0.03 mm after finishing.
Other limitations include the use of bovine instead of human teeth and the variable morphology and size of the bovine teeth. Natural teeth tend to have various morphologies as well, so this was considered a better representation of the in vivo situation. Besides, standardizing the amount of bonding material and keeping fixed distance between composite bonding spots would minimize the effect of this variability.

| STRENGTH OF THE STUDY
The strength of the study can be seen in that it is the first study which investigates the performance of PEEK wires with different surface treatments as a fixed retainer wire and compares its performance to other metallic counterparts.

| CONCLUSIONS
Within the limitations of this in-vitro study regarding the efficacy of PEEK wire as a retainer after orthodontic treatment, it can be concluded that the 0.8 mm round wire-form has comparable performance-in terms of debonding and pull out forces-to conventional retainers when bonded with 4 mm composite bonding spots; using air-abrasion for 10 s at 3.5 MPa provided adequate adhesion of the composite to the wire, and conditioning with the adhesive system Visio.Link may provide no further clinical benefit.
The use of PEEK wire over other metallic wires is supported for future in vivo study to confirm these outcomes.