The search generated 797 reviews in PubMed, 114 in Web of Science, 97 in Cochrane Library, and 319 in Google Scholar (Fig. 1). The following were selected after the title and abstract screening: 18 were selected from PubMed, 23 from Web of Science, 6 from Cochrane Library, 2 from Google Scholar, and 2 from hand searching the reference list of the selected article. After the duplicates were removed, 30 articles remained for full-text analysis. Twenty-four were eliminated after full-text analysis, and six articles remained for data extraction (Table 2). The risk of bias of the selected studies scored mostly “definitely low risk of bias” or “probably high risk of bias” (Table 1).

Most of the selected six studies reported that new bone formation and other histomorphometric outcomes were not statistically different between control and test sites. Only one study reported a significant increase in new bone formation between BMAC + bovine bone graft test sites compared to bovine bone graft controls. Another study reported histologic outcomes of BMAC-grafted sites to produce as much new bone as the traditional laboratory-based method. Most studies also reported that all implants placed in both test and control sites were osseointegrated, successfully loaded, and in function.

The reported ranges of vital mineralized tissue for the included studies for the BMAC groups were 34.63–55.15% compared to 27.30% for control groups [7, 12]. For vital mineralized bone, these studies reported variable statistical significance. At 3–4 months, new bone formation for BMAC groups with controls [2, 14, 15] using no BMAC was 7.4–12.6% and for the control groups was 9.45–14.3%. At 6 months, new bone formation for BMAC groups [2, 15] was 13.5–14.12% and for control groups was 10.41–13.9%. For new bone formation, most studies reported no significant difference between the test and control and between the 3 and 6 months histologic evaluation.