Misfit simulation on implant-supported prostheses with different combinations of engaging and nonengaging titanium bases: Part 3: A radiographic evaluation

Statement of problemThe fit of implant-supported prostheses plays an important role in their mechanical and biological stability. Clinically, the prosthetic fit is typically assessed radiographically, but this method relies on the operator's subjective evaluation. Whether available digital tools could optimize the evaluation of the prosthetic fit is uncertain.PurposeThe purpose of this in vitro study was to evaluate the influence of an image processing program on the radiographic detection of discrepancies in the active and passive fit of implant-supported prostheses. Two-implant-supported screw-retained prostheses were analyzed by simulating the vertical and horizontal misfits of 3 different implant abutment configurations.Material and methodsSeven casts were fabricated using 2 internal-connection titanium implants: 1 control; 3 with vertical (V) misfit of 50 µm, 100 µm, 150 µm; and 3 with horizontal (H) misfit of 35 µm, 70 µm, 100 µm. Thirty bar-shaped zirconia frameworks were fabricated and divided into 3 groups (n=10) according to their attachment to 2 engaging (E-E), 2 nonengaging (NE-NE), and engaging and nonengaging (E-NE) titanium bases. Digital parallel periapical radiographs were made of each specimen in the passive and active fit situation on each cast (1-screw test), except for the E-E specimens, which were only seated on the control, H35, and H70 casts because the fit on the remaining casts was poor. The mean gray value (MGV) was measured at the chosen regions of interest on the second implant (side B) using the ImageJ software program. Differences in the MGV measurements between the passive and active conditions were tested using a t test (α=.05) and compared the different misfit levels using analysis of variance (1-way ANOVA), followed by the Tukey HSD test (α=.05).ResultsThe highest values for the differences between passive and active fit were found for the V150 and H100 misfit simulations (P<.05). Statistical differences between the MGVs were found with some exceptions: the smallest simulated misfits (H35 and V50) revealed statistically significant MGV differences from the highest simulated misfits (V150, H100) and from the H70 in the groups where an engaging component was present (P>.05). In the horizontal misfit group of NE-NE abutment configuration, H70 revealed no significant difference from the control group cast (P>.05).ConclusionsMeasuring MGV differences between passive and active fit could be a promising alternative for detecting 70- to 150-µm gaps in the implant-abutment connection that result from the misfit. However, the procedure was not adequate for detecting <50 µm gaps, cannot be uniformly applied to all types of implant-abutment connections, and requires 2 exposures to X-radiation.