Alterations in Bone Quality After Socket Preservation with Grafting Materials: A Systematic Review

Purpose: While the ability of various grafting materials to preserve extraction socket morphology has been adequately reviewed, the quality of the grafted bone in the socket is not as well understood. This systematic review aimed to compare the proportion of vital bone and connective tissue between grafted and naturally healed sockets. Materials and Methods: An electronic search of five databases (from 1965 to November 2011) and a hand search of peer-reviewed journals for relevant articles were performed. Human clinical trials that compared histologic components of soft and hard tissues in augmented sockets and naturally healed sites, with at least five samples per group, were included. Results: Eight studies (five randomized controlled trials and three controlled clinical trials) were included. The mean percentages of vital bone and connective tissue in natural healing sockets were 38.5% ± 13.4% and 58.3% ± 10.6%, respectively. Limited evidence (one to two articles for each material) implied that vital bone fraction was not different with demineralized allografts and autografts and increased by 6.2% to 23.5% with alloplasts in comparison to nongrafted sites. Four studies investigating the effect of xenografts were available, with equivocal results. The difference in the mean percentage of vital bone ranged from -22.2% (decrease) to 9.8% (increase). Connective tissue content decreased with the use of the aforementioned bone substitutes. Considerable residual hydroxyapatite and xenograft particles (15% to 36%) remained at a mean of 5.6 months after socket augmentation procedures. Conclusions: Based on a limited number of prospective comparative studies, the use of grafting materials for socket augmentation might change the proportion of vital bone in comparison to sockets allowed to heal without grafting. Whether these changes in bone quality will influence implant success and peri-implant tissue stability remains unknown.