OBJECTIVE To evaluate the efficacy of manipulation therapy in repairing thin-layer and thick-layer articular cartilage defects in rabbits. METHODS Twenty-four rabbits were randomly divided into manipulation group and control group. Chondral defects in the center of condyles medialis femoris (punch diameter 2.0 mm, thin-layer defect, without damaging the subchondral bone) and the center of facies patellar femoris (punch diameter 3.0 mm, thick-layer defect, without damaging the subchondral bone) were made in both side knee joints of the rabbits. Four weeks after chondral defect, the rabbits in the manipulation group were treated with manipulation therapy, while the rabbits in the control group didn't take any therapeutic measure. The rabbits were sacrificed after 4-, 8- and 12-week manipulation treatment to obtain the distal femur. The appearance of the chondral defect was first assessed, and then regenerated tissues were stained respectively with hematoxylin and eosin (HE), safranin O and immunohistochemical methods to observe the changes in pathology. The regenerated tissue O'Driscoll score system was also used to evaluate the efficacy of manipulation therapy. RESULTS After manipulation treatment, although the thin-layer articular cartilage defects of the control group were partly repaired, a sharp gap between the regenerated tissue and the normal cartilage was visible. The range of the thick-layer articular cartilage defect in the control group became smaller, and the regenerated tissue only consisted of fiber tissue. On the contrary, both the thin-layer defects and the thick-layer defects in the manipulation group were well repaired. The appearance of the regenerated tissues was close to the normal cartilage. The gap between regenerated tissue and normal tissue was disappeared and the regenerated tissue was well integrated with around tissues. HE staining showed that the regenerated tissues in the manipulation group were similar to hyaline cartilage. Safranin O staining showed that matrix stain of the regenerated tissues in the manipulation group was similar to the normal cartilage, but no stain was observed in the control group. The type II collagen immunohistochemical staining showed that the color of the regenerated tissues was a little darker than that of the normal cartilage in the two kinds of defects in the manipulation group, and the two kinds of defects in the control group were little stained. The O'Driscoll cartilage scores of the two kinds of defects in the manipulation group were higher than those in the control group at different time points (P
