Introduction
Moving teeth rapidly and avoiding posttreatment relapse are fundamental goals of orthodontic treatment. In-vitro and animal studies suggest that the human hormone relaxin might increase the rate of movement and the stability through its effect on the periodontal ligament. The purpose of this study was to compare relaxin and a placebo with regard to tooth movement and stability in human subjects.
Methods
A single-center, blinded, placebo-controlled, randomized clinical trial was used to examine the effect of relaxin on tooth movement and stability. Forty subjects were randomized 1:1 and received weekly injections of 50 μg of relaxin or a placebo for 8 weeks. Aligners programmed to move a target tooth 2 mm during treatment were dispensed at weeks 0, 2, 4, and 6. Movement was measured weekly on polyvinyl siloxane impressions that were scanned and digitized. The subjects were followed through week 12 to assess relapse.
Results
Tooth movement over the 8-week treatment period did not differ by treatment group ( P = 0.995). By using an intent-to-treat analysis, we found that the mean tooth movement for both groups was 0.83 mm (SE, 0.08 for relaxin and 0.09 for the placebo). Relapse from weeks 8 to 12 was the same in both groups ( P = 0.986), and the mean was −0.75 (SE, 0.07 for relaxin and 0.08 for theplacebo).
Conclusions
No differences in tooth movement over 8 weeks of treatment or relapse at 4 weeks posttreatment were detected when comparing subjects who received weekly injections of relaxin with those who received a placebo. In both groups, an average of less than half of the programmed tooth movement was obtained after 8 weeks of treatment. The local doses of relaxin might have been too low to affect tooth movement or short-term relapse.
Relaxin is a naturally occurring hormone that was discovered in 1926. This peptide was initially associated with pregnancy, specifically relaxing the pubic ligaments and softening and enlarging the opening of the uterus. More recently, its role in maternal hemodynamic and renal adjustments required during pregnancy has been of interest, with implications for treatment of heart failure. In both sexes, relaxin impacts other physiologic functions, including collagen turnover, angiogenesis, and antifibrosis.
Because of its role in remodeling soft tissues, the use of relaxin in models of tooth movement has been studied. Tooth movement, in response to force, involves changes to the periodontal ligament, gingival tissue, and alveolar bone. With tooth movement, the gingival response is to increase collagen formation to resist movement. The gingival “memory” also plays a role in relapse after treatment. In-vitro models have been used to examine the effect of relaxin on stretched human periodontal ligament cells. Takano et al found that relaxin modulates collagen metabolism, affecting the release and expression of type I collagen and matrix metalloproteinase-1. This suggests that the use of relaxin might be beneficial in preventing relapse after orthodontic treatment. Henneman et al examined matrix metalloproteinase expression in human periodontal ligament cells and found no increase in total matrix metalloproteinase production, although there was a dose-dependent increase in matrix metalloproteinase-2 production. Stewart et al found that relaxin receptors were localized in the canine supracrestal periodontal ligament cells used to study the effect on rotational orthodontic relapse.
Animal studies have also examined the use of relaxin in tooth movement. Tooth relapse was studied in dogs, by using control, fiberotomy, systemic plus gingival relaxin, and gingival relaxin injection treatment groups. There was a significant difference in relapse between the control and the fiberotomy groups (33% vs 18% mean relapse, respectively) but not in the groups treated with relaxin. Because of outliers, when the medians of the relaxin groups were assessed, the systemic plus gingival group was similar to the control group, whereas the gingival injection group was similar to the fiberotomy group, suggesting that gingival application of relaxin attenuated relapse. It was found that animals receiving systemic relaxin developed neutralizing antibodies to relaxin; this might explain the lack of efficacy in this group.
Three groups of young rats received human relaxin through a mini-osmotic pump, subcutaneous injection, or placebo (pump), and tooth movement was assessed. Because of large individual variations and small sample sizes, statistically significant differences were not found for sagittal tooth movement over 14 days. Both relaxin-treated groups had increased tooth movement at day 3 compared with the controls, and postmortem measurements of molar lengths and intermolar spaces differed significantly for the control and relaxin groups. A second rat study did not detect differences in tooth movement; it compared relaxin administered by mini-osmotic pump vs a placebo. However, additional analyses suggest that relaxin did affect the structure of the periodontal ligament, since (1) the timing of tooth movement differed, (2) the fractal analysis suggested greater variability at earlier times for the rats treated with relaxin, and (3) increased molar mobility was noted in the relaxin group.
To further understand the potential role of relaxin in tooth movement, we examined the rate of orthodontic tooth movement and subsequent relapse in humans. Since relaxin acts on the breakdown of collagen and produces changes in the periodontal ligament, resistance to orthodontic forces by both bone and soft tissue might be reduced, resulting in increased rates and amounts of tooth movement. Also, because of changes in the periodontal ligament, relapse upon force removal might be minimized.
Material and methods
This study was a single-center, randomized, placebo-controlled evaluation of tooth movement in patients needing minor incisor alignment. The subjects were healthy adults who were scheduled to receive orthodontic treatment. They resided in north-central Florida, and the study was conducted between April and November 2005. The Institutional Review Board for the Protection of Human Subjects at the University of Florida approved the study.
To determine enrollment, 2 screenings were performed. The purpose of the initial screening was to identify potential subjects with malocclusions needing minor incisor alignment of at least the maxillary incisors and to eliminate those with medical conditions or intraoral problems. To that end, an intraoral clinical examination was performed. This included identification and assessment of pulp vitality of the target tooth, either the right or left maxillary central incisor, and the 2 immediately adjacent teeth by using cold thermal stimulation, tissue health and gingival recession by visual inspection, papillary bleeding score, and periodontal probe (Florida Probe, Gainesville, Fla) to evaluate pocket depth. Subjects who were eligible underwent a second screening that included (1) a physical examination, including measurements of height weight, and an electrocardiogram; (2) vital signs, including blood pressure, heart rate, respiratory rate, and temperature; (3) blood collection for hematology and fasting serum chemistries; (4) a serum sample collected for analysis of anti-relaxin antibodies and serum pregnancy tests on female subjects; (5) urine collection for urinalysis; (6) concomitant medications; (7) diagnostic quality maxillary anterior periapical x-rays as the standard of care and to establish the crown-to-root ratio; (8) impressions with polyvinyl siloxane for preparation of Invisalign (Align Technology, Santa Clara, Calif) appliances (the impressions were sent to Align after confirmation of eligibility and randomization); and (9) intraoral and extraoral photographs, as well as panoramic radiographs and lateral cephalometric x-rays. After a review of all information and laboratory results to confirm eligibility, the subjects were enrolled into the study and randomized to the relaxin or the placebo group and assigned a unique study number. The subjects were randomized in 2 cohorts. The initial single-blind cohort consisted of 12 subjects; this was done for safety considerations. The second double-blind cohort of 28 subjects began after all subjects in the initial cohort had completed at least 4 weeks of treatment. Both cohorts were randomized in blocks of 4, with half assigned to each treatment. The study biostatistician generated the randomization scheme, and the treatments were dispensed by the investigational pharmacy.
Once each subject was accepted into the trial and the polyvinyl siloxane impression had been obtained, the right or left maxillary central incisor was selected as the target tooth. The selection was based on the target tooth not being blocked out by adjacent teeth to allow anteroposterior movement of 2 mm. A series of 4 maxillary aligners was programmed with 0.5 mm of anteroposterior movement for the selected central incisor. Aligners were dispensed at weeks 0, 2, 4, and 6. If the subject could not tolerate the next aligner, he or she kept the current aligner. Only crown tipping of 1 central incisor in the anteroposterior dimension was programmed (ie, no intrusion, extrusion, or rotation was attempted).
All subjects received 8 weekly treatments of relaxin or the placebo. The doses were determined based on animal and human studies. Although relaxin levels vary during pregnancy, concentrations are generally in the range of 0.97 to 1.01 ng per milliliter. Studies have indicated that circulating levels of 1 to 20 ng per milliliter are effective in mediating changes in the extracellular matrix deposition and affecting renal and cardio hemodynamics. A relapse study carried out in dogs used 4 gingival injections of 125 μg each, with the animals treated on days 55 and 60 for a total exposure of 1000 μg. Since it was impossible to predict the local levels of tissue concentration at the gingiva and the periodontal ligament, our dose of 50 μg per treatment ensured an initial high local concentration, which decreased as the relaxin diffused from the injection site. Each treatment consisted of 2 injections of 0.1 mL of the study drug injected (infiltration) on either side of the targeted tooth. Each injection of the study drug was administered to the gingival tissue to a depth of approximately 2 mm by using a 0.3-mL syringe and a 31-gauge needle. Approximately half of each injection was delivered to the lingual side of the tooth, with the remainder delivered to the facial side on withdrawal. The total dose of relaxin at each treatment visit was 50 μg (0.2 mL), or 25 μg per injection; the maximum total cumulative dose for a subject completing the 8 weekly injections was 400 μg. The subjects assigned to the placebo group received 2 injections (each, 0.1 mL) of the vehicle. The study treatment period was 8 weeks (weeks 1-8), followed by a 4-week posttreatment evaluation of retention of the anteroposterior movement of the target tooth (weeks 9-12), and then a final visit at week 32 for safety (6 months after the treatment).
For the safety assessment, baseline (predose) and 2-hour postinjection blood samples were drawn at the initial trial visit. The serum relaxin levels were determined.
The weekly polyvinyl siloxane impressions were sent to Align Technology, where they were laser scanned to create 3-dimensional digital images of tooth positions. Treat III, proprietary software created by Align Technology, was used the digital study models for 3-dimensional computer analysis of tooth movement through superimposition techniques. Before superimposition of time points week 0 and subsequent time points, the subject’s surface anatomy of each tooth at the 2 time points was matched, excluding teeth adjacent to the tooth being moved. Matching deviations for each tooth in the arch were given by average deviations (millimeters), maximum deviations (millimeters), and coverage percentages. Since tooth anatomy is stable, any deviation between the 2 time points would indicate some degree of impression inaccuracy or improper placement of the tooth axis for 1 time point. If matching deviation was negligible, the best-fit superimposition of the digital study models between time points was performed by using reference teeth and a reference stage. The software automatically selected the stable reference teeth by a statistical filtering process. The reference stage corresponded to the aligner number that the subject was wearing at the time of the study model impressions. This allowed the software to compare the amount of tooth movement achieved with the amount attempted relative to the reference stage. Raw data of tooth movement obtained from the superimposition was exported into a spreadsheet. Quantitative treatment changes for each tooth were described as changes in the x, y, and z coordinates for translation, with the Euclidean distance of movement of the centroid of the crown determined.
Statistical analysis
Demographic characteristics were compared for the placebo and relaxin groups by using standard statistical tests. An intent-to-treat analysis, with the last observation carried forward, was used to evaluate anteroposterior movement of the target tooth from week 0 through week 8 (movement), and week 8 through week 12 (retention), comparing relaxin with the placebo. All subjects who received at least 1 treatment and aligner were included in the movement comparison, and the relapse comparison included all subjects wearing an aligner at week 8. A per-protocol analysis was also performed. To qualify for this analysis, a subject must have received at least 7 of the 8 treatments. For the primary outcomes, 2-sided, 2-sample t tests were used to compare treatment groups, with a P value less than 0.05 considered statistically significant. We also examined the ability to progress to the next set of aligners over the 8-week treatment period. Chi-square tests (or Fisher exact tests for small expected cell sizes) were used to determine treatment differences. In addition, linear mixed model analyses were used to examine the pattern of change over the 8-week treatment period. This analysis considers each weekly movement, accounting for correlation in a subject, and is not sensitive to randomly occurring missing values. Exploratory analysis examined the impact of sex and age on tooth movement during the 8 weeks of aligner use, by using the per-protocol subjects.
Results
A total of 40 subjects were enrolled in this study; however, 1 subject declined to participate before treatment. Demographic characteristics and comparisons of the treatment groups are shown in Table I . More women (72%) than men (28%) were enrolled, with an overall mean age of 26.9 years (SD, 5.1; range, 18.6-40.5 years). In the relaxin group, the target tooth was equally distributed between the maxillary right and left central incisors, whereas in the placebo group the maxillary left incisor was used more than the right one ( Table II ). Overall, 90% of the subjects completed 7 of the 8 doses of either relaxin or placebo ( Table III ). There were no statistically significant differences for these characteristics between the placebo and relaxin groups. No health issues were determined from physicals, electrocardiograms, urinalysis, or hematology attributed to relaxin. At the safety assessment, serum relaxin levels increased in all relaxin injected subjects (predose median, <18.8 pg/mL, range, <18.8-383 pg/mL; postdose median, 1016 pg/mL, range, 598-1532 pg/mL), whereas only minor increases were observed in 3 placebo subjects (predose median, <18.8 pg/mL; range, <18.8-307 pg/mL; postdose median, <18.8 pg/mL; range, <18.8-376 pg/mL). No anti-relaxin antibodies were found in any subject treated with the placebo or relaxin at any time point. For the tooth movement portion of the study (weeks 0-8), the intent-to-treat analysis included 20 relaxin and 19 control subjects; the per-protocol analysis included 18 relaxin and 17 control subjects. For the retention portion of the study (weeks 8-12), the intent-to-treat analysis included 19 relaxin and 18 control subjects; the per-protocol analysis included 18 relaxin and 17 control subjects.
| Variable | Parameter | Relaxin | Placebo | P value ∗ |
|---|---|---|---|---|
| n = 20 | n = 19 | |||
| Sex (n) | Male | 4 (20.0%) | 7 (36.8%) | 0.24 |
| Female | 16 (80.0%) | 12 (63.2%) | ||
| Age (y) | Mean | 26.2 | 27.7 | 0.35 |
| SD | 4.5 | 5.8 | ||
| Minimum | 19.2 | 18.5 | ||
| Maximum | 37.7 | 40.5 | ||
| Race (n) | White | 14 (70.0%) | 16 (84.2%) | 0.45 |
| Black | 4 (20.0%) | 1 (5.3%) | ||
| Asian | 1 (5.0%) | 0 (0.0%) | ||
| American Indian | 0 (0.0%) | 0 (0.0%) | ||
| Hispanic | 1 (5.0%) | 2 (10.5%) |
∗ The chi-square test was used for sex, the 2-sample t test for age, and the Fisher exact test (white vs other races) for race.
