Contrary to conventional hollow-screw implants (only marginal gap), a problem of the expandable implant is the presence of gaps down to the apical region. Former microbial assessment of different implant-abutment interfaces displayed that none of the marginal connections had the capacity to prevent microbial leakage [63,64,65]. Therefore, an apical microleakage (comparable to distractable implants and endodontically treated teeth) might be a disadvantage of the evolved implant system [66, 67]. However, according to the manufacturer’s information, a microbiological study revealed no microbial leakage through the expanded implants. Over the follow-up period, we equally did not observe any inflammatory signs in the apical region, neither clinically nor radiologically (Figs. 4b and 5c–d). Nevertheless, this aspect should be analysed under mechanical loading in vitro. Based on an earlier animal histologic study [47], as well as a clinical up-to-40-month study [46], which referred to comparable apically expandable implants, authors did not report any periapical inflammatory complications. To eliminate the potential risk of deep intrabony microleakage, it is questionable whether equal biomechanical stability values can be achieved only by the macrothread design avoiding any deep microgaps.

In the present study, the crestal bone changes under loading in the first year exceeded that of the second year. Moreover, the differences between the maxilla and mandible in each year were not statistically significant, which only partially agrees with previous findings in the literature [7, 58]. Besides microbiological conditions, there are several biomechanical aspects which influence maintenance of periimplant crestal bone. Conical and parallel surfaces of the implant-abutment connection (internal hexagon) provide rotational stability and reduce microgaps and micromovement [68]. Another important factor is the thickness of the implant shoulder [69], which might be a weak point in the design of a short implant due to elastic deformity under extra-axial loading. This fact might be the reason for non-inflammatory periimplant crestal bone loss. We addressed this aspect by splitting adjacent implants wherever possible [50, 51]. According to Brenner and co-authors [30] as well as Pommer and co-authors [50], the following prosthodontic factors are to be considered to avoid screw loosening, component fracture, loss of marginal bone or even loss of osseointegration: crown-to-implant ratio (extra-axial loading), cantilever length, status of opposing dentition, splinting of adjacent implants, occlusal surface relief and dimensions.