“…Among included studies, fourteen studies used cranial bone defect to conduct their stud [ 35 – 38 , 42 – 45 , 47 – 52 ], while only four studies chose the lower limb bone defect model (tibia/femoral bone defect) [ 39 – 41 , 46 ]. Bone defect model size was equal or larger than 5 mm in eleven studies [ 35 – 37 , 42 , 44 , 45 , 47 – 50 , 52 ], while it was lower than 5 mm in seven studies [ 38 – 41 , 43 , 46 , 51 ]. It was noteworthy that six studies did not provide their sample size [ 36 , 38 , 42 , 48 , 49 , 51 ], which contributed to these studies could not include in further meta-analysis.…”