The IT, RT, and ISQ values revealed significant differences among the implants (Table 2).

The IT and RT values of 12S were not significantly different compared with published data (IT, 13.13 ± 1.763 N cm; RT, 12.37 ± 1.746 N cm) (Student t test, df = 9, t = 2.91, p < .017). Compared with 12S, the IT and RT values of 06D and 06S were significantly different (147% and 150%, and 163% and 173%, respectively). The results were analyzed by a statistician who was blinded to the details of the study. The IQ of 06S (22.30 ± 1.68) was significantly higher compared with those of D06 (20.38 ± 1.62), and S12 (13.67 ± 1.78) (Student t test, df = 9, t = 6.06, p < .001) was significantly higher compared with those of D06 (17.45 ± 1.22) (Student t test, df = 9, t = 5.41, p < .001) and S12 (11.69 ± 1.01). The ISQ values of 06D (53.77 ± 2.59) and 06S (55.66 ± 1.55) were significantly higher compared with that of 12S (51.40 ± 2.81). There was no significant difference between 06D and 06S (ANOVA, F (2, 27) = 7.24, p < .001) (Fig. 2). The RT values of the three implants were significantly lower compared with the IT values (Fig. 3).

The 12S and 06D implants reached maximum torque twice as fast as the 06S implant

The IT values of the three implants increased approximately linearly with time (Fig. 4). The periodicities of the implants were consistent with rotation. Similar maximum torque values were reached by 12S and 06D, which were 2-fold shorter compared with that of 06S, consistent with the lengths of the leads. Maximum removal torque values were reached within 2 s or after a half-turn (Fig. 4). Removal torque-time curves were linear with periodic waves. The removal times were comparable between 12S and 06D, which were 2-fold shorter compared with that of 06S.

The 06D implant damages bone to a greater extent compared with the 06S or 12S implant

Microscopic observations of the contact interfaces of the artificial bone and implant body revealed minor damage to artificial bone tissue, characterized by debris and voids in the artificial bone that adhered to the implant (Fig. 5a-c). The 06D implant caused the most severe damage, followed by 06S and 12S. In the artificial bone tissue adjacent to 06D, there were numerous voids and abundant debris. Morphometrical analysis revealed that the number of debris particles attached to 06D (73 particles) (Fig. 5b) was higher compared with those of 12S (55 particles) (Fig. 5a) or 06S (52 particles) (Fig. 5c). Although the numbers of debris particles were comparable between 12S and 06S, their sizes were greater in 06S (535.8 μm²) compared with those of 12S (305.4 μm²) (Fig. 6b). Individual debris particles were classified as follows: small (< 1000 μm²), medium (1000–10000 μm²), and large (≥ 10000 μm²) (Fig. 6a-c). Most debris particles (61%) of 12S were small, with no detectable large debris particles. In contrast, most debris particles associated with 06D and 06S were medium (55% and 65%, respectively) or large (12% and 5.6%, respectively).