The results from the histomorphometric measurements are depicted in Table 2. Bone biopsies were composed by 21.37% (SD 7.36) of residual CCXBB, 26.90% (SD 12.21) of mineralized vital bone (MVB), 47.13% (SD 19.15) of non-mineralized tissue and 0.92% of DBBM (Fig. 5b). Biopsies from patients who lost their implants had a statistical significant lower amount of MVB (p = 0.01^u) and a statistical significant larger proportion of connective tissue (p = 0.02^t) (Table 4). Furthermore, although no statistically significant correlation was observed between presence of soft tissue dehiscence and specific histomorphological outcomes, a tendency towards a low amount of new bone was observed in the specimens from patients where the bone graft had been exposed (p = 0.06).

Results from the immune-histochemical analysis are presented in Table 3. A large number of OPN-positive cells (mesenchymal, hematopoietic cells and osteoblast) were observed in most of the tissue samples (Fig. 6a). Similarly, TRAP positive multinuclear cells (osteoclasts) were also observed mainly in contact with the residual CCXBB (Fig. 6b). More limited amounts of OSC-positive cells (mature osteoclast) were observed (Fig. 6c) whereas ALP-positive cells (osteoblast) were mainly detected on the surface of the newly formed woven bone and in proximity of vascular units (Fig. 6d). The newly formed bone in close contact with the CCXBB remnants showed signs of modelling and remodelling. When the correlation between the immunohistochemical results and implant loss was investigated, a statistically significant correlation between implant loss and number of OSC positive cells was observed (2 versus 8.78% p = 0.02^u) (Table 4).