Objectives:It has often been suggested that bone-patellar tendon-bone (BTB) grafts heal more quickly to graft tunnels than soft-tissue grafts after ACL reconstruction, due to rapid bone-to-bone healing. Little is known, however, about the motion of the grafts under functional, dynamic loads, or the actual time course for graft-tunnel healing in humans. The goal of this study was to assess the motion of both hamstrings (HS) and BTB grafts during gait at two time points (6 and 52 weeks) after surgery. It was hypothesized that there would be greater graft-tunnel motion in HS than BTB grafts, and that graft motion would be less at 1 year when compared with 6 weeks.Methods:Twelve human subjects underwent anatomic single-bundle ACL reconstruction using either hamstrings (HS) or bone-patellar tendon-bone (BTB) autografts (6 in each group) and extra-cortical fixation. Six 0.8 mm tantalum beads were embedded into the grafts prior to implantation, with pairs of beads located within each bone tunnel and in the mid-substance region. At six weeks and one year after surgery, dynamic stereo X-ray (DSX) images were collected at 100 frames/s while subjects walked on an instrumented treadmill (1.3 m/s). 3D femur and tibia bone models were generated from CT scans. Tibiofemoral kinematics and graft bead positions were analyzed by combining the 3D models with DSX data. Graft-tunnel motion was assessed by determining the maximum range of motion of the most proximal femoral tunnel and most distal tibial tunnel beads along the direction of the bone tunnels during the stance phase of gait. Effects of graft type and time were assessed with a 2-factor ANOVA, with alpha = 0.05.Results:Femoral tunnel motion (mean±s.d.) at 6 weeks was 2.47±0.96 mm (range 0.95-3.44 mm) for BTB and 1.51±0.68 mm (range 0.88-2.41 mm) for HS grafts, decreasing slightly to 2.09±0.84 mm for BTB and 1.37±0.41 for HS grafts by one year. Tibial tunnel motion was 1.45±0.57 mm (range 0.74-2.44 mm) for BTB and 1.43±0.59 mm (range 0.89-2.30 mm) for HS grafts at 6 weeks, and changed little by one year (1.58±0.46 mm for BTB and 1.52±0.57 for HS grafts). Across all time points, femoral graft-tunnel motion was significantly greater for BTB than HS grafts (p=0.004). In the tibial tunnel, there were no significant differences in motion between BTB and HS at 6 weeks or 1 year. For both graft types, there was no significant change in motion between 6 weeks and 1 year.Conclusion:Contrary to the study hypothesis, there was a greater motion relative to the femoral tunnel wall in BTB than HS grafts during walking. Significant graft-tunnel motion also persisted one year after reconstruction. These data suggest that the graft-to-tunnel healing process may occur more slowly then widely believed, especially for BTB grafts. However, all subjects were asymptomatic one year after surgery, and the implications of persistent, small (1-3 mm) movements of the grafts are unknown. These results, however, may have significant implications for graft choice, rehabilitation strategies and timing for return to sports.