Background: Peri-implant marginal bone loss is influenced by the interaction between tissues and the implant–abutment complex. The implant–abutment connection geometry is considered to be one of the factors that most affects peri-implant bone remodelling.
Keywords: implant connection, implant microgeometry, implant stability, implant stability quotient value, marginal bone levels, neural network, randomised controlled trial, shoulderless preparation, zirconia abutment
The authors declare no conflicts of interest related to this study.
Purpose: The primary purpose was to compare the clinical and radiographical differences between implants sharing the same macrogeometry but with two different connections. The secondary aims were to measure implant success and survival rate, primary and secondary stability, and the correlation between changes in marginal bone level and clinical variables. Additionally, a neural network was developed and tested to anticipate the impact of the insertion torque curve on marginal bone loss.
Materials and methods: Patients requiring at least two implants in the posterior region were randomly divided into two groups. The implants presented the same micro- and macrotopography with different internal connections, conical standard (CS) and internal hex (IH). Upon implant surgery (T0), insertion torque, implant stability (implant stability quotient values were recorded by resonance frequency analysis), soft tissue height and the amount of keratinised gingiva were assessed. Stability was remeasured at the time of prosthetic connection (stage-two surgery) using a one-abutment one-time protocol and a fully digital workflow. At 6 months and 1 year after implant loading, periodontal parameters were assessed and periapical radiographs were taken. To study the differences between the two groups and the different variables, paired t test and generalised estimating equations models were adopted. Cluster analysis was used to assess the correlation between torque insertion/clinical profiles and changes in marginal bone level.
Results: A total of 33 patients (17 men, 16 women, mean age 67.4 ± 14.5 years) were included in the study. No dropouts were reported. Fifty-three implants (26 CS and 27 IH) were inserted in the maxilla, and 15 (8 CS and 7 IH) in the mandible. No implants failed. Marginal bone loss at 6 months after prosthetic loading was 0.33 ± 0.34 mm for CS and 0.43 ± 0.37 mm for IH (P = 0.125), and after 1 year was 0.48 ± 0.18 mm for CS and 0.57 ± 0.24 mm for IH. A statistically significant difference between the implant stabilty quotient values for the test and control groups was demonstrated at T0 (P = 0.03) and at stage-two surgery (P = 0.000122). The generalised estimating equations model showed that soft tissue height (P = 0.012), keratinised gingiva (P = 0.05) and insertion torque (P = 0.042) had a significant effect on marginal bone loss, while the other variables did not play a statistically significant role. The neural network showed good sensitivity, accuracy, precision and specificity.
Conclusions: The present research showed that different implant–abutment connections with the same implant macrogeometry have a significant effect on marginal bone loss. Better outcomes were observed in the CS group compared to the IH group. Marginal bone loss was found to be influenced by different individual and clinical factors.