Marginal bone-level alterations of loaded zirconia and titanium dental implants: an experimental study in the dog mandible

ObjectivesThe aim was to test whether or not the marginal bone-level alterations of loaded zirconia implants are similar to the bone-level alterations of a grade 4 titanium one-piece dental implant. Materials and methodsIn six dogs, all premolars and the first molars were extracted in the mandible. Four months later, three zirconia implants (BPI, VC, ZD) and a control titanium one-piece (STM) implant were randomly placed in each hemimandible and left for transmucosal healing (baseline). Six months later, CAD/CAM crowns were cemented. Sacrifice was scheduled at 6-month postloading. Digital X-rays were taken at implant placement, crowns insertion, and sacrifice. Marginal bone-level alterations were calculated, and intra- and intergroup comparisons performed adjusted by confounding factors. ResultsImplants were successfully placed. Until crown insertion, two implants were fractured (one VC, one ZD). At sacrifice, 5 more implants were (partly) fractured (one BPI, four ZD), and one lost osseointegration (VC). No decementation of crowns occurred. All implant systems demonstrated a statistically significant (except VC) loss of marginal bone between baseline and crown insertion ranging from 0.29mm (VC; P=0.116) to 0.80mm (ZD; P=0.013). The estimated marginal bone loss between baseline and 6months of loading ranged between 0.19mm (BPI) and 1.11mm (VC), being statistically significant for STM and VC only (P<0.05). The changes in marginal bone levels were statistically significantly different between zirconia implants and control implants (STM vs. BPI P=0.007; vs. VC P=0.001; vs. ZD P=0.011). ConclusionsZirconia implants were more prone to fracture prior to and after loading with implant-supported crowns compared to titanium implants. Individual differences and variability in the extent of the bone-level changes during the 12-month study period were found between the different implant types and materials.