Methods: Twelve orthodontic patients (2 males, 10 females, aged 11.6 to25.0 years) were selected. Enrollment criteria contain moderate crowding malocclusion, extraction of upper first premolars, and moderate anchorage. They were divided into TPA groups and control group. The upper study casts were collected at the initial treatment and at the end of aligning and leveling stage. Next, the dental casts was scanned by LPX-1200 3-dimensional laser scanner ( Roland, Japan ), and analyzed by 3D reverse modelling software program (Rapidform 2006®, South Korea). The coordinate system was constructed including the X-Z plane (the occlusal plane), X – Y plane (sagittal plane) and the Y – Z plane (frontal plane). Positive values along the X -. Y -axis and Z-axis indicated mesial, intrusive and left movements. Then the displacement of mesiobuccal cusps (MB)，mesiopalatal cusps (MP), distobuccal cusps (DB) and distopalatal cusps (DP) of the upper first molars were evaluated by the Rapidform based on the surface-to-surface matching function of the autoscan system on the palatal vault region of the casts.
Results: Vertically the displacements of MB, DB and DP (TPA: 0.59,0.55 and 0.57mm & control group:-0.42,-0.30 and -0.35mm) reached statistically significant difference , and anteroposteriorly and transversally MB, MP, DB and DP did not reach by SPSS13.0.
Conclusion: TPA was an effective anchorage method vertically during the alignment and leveling stage. This imply TPA may be the effective control for the cases of the skeletal Ⅱ with an increased mandible plane. However, there seems no differences between groups with or without TPA anteroposteriorly and transversally. Therefore, other measure should be taken in order to attain ideal treatment goal for the given case.