Gadolinium enhanced MRI assessment of bone-patellar tendon-bone graft harvest on patellar vascularity

Abstract

Objectives: Bone-patellar tendon-bone (BPTB) autograft remains a favored graft source for anterior cruciate ligament (ACL) reconstruction despite problems related to donor-site morbidity. Patellar devascularization has been proposed as a source of anterior knee pain following vascular disruption from traumatic injury (fracture) or surgical procedures involving the patella (total knee arthroplasty); however, no study has investigated the effect of BPTB harvest on patellar vascularity. Recent anatomic studies have suggested that the dominant arterial supply enters the patella through the inferior pole. We hypothesized that BPTB harvest can significantly diminish patellar vascularity following graft harvest. Methods: Nine matched pair cadaveric knee specimens (mean age 47.4 years) were dissected and cannulated at the superficial femoral, anterior tibialis, and posterior tibialis arteries. A single knee was randomly selected to undergo bone graft harvest. The contralateral knee was left intact to serve as a control. Gadolinium (Gd-DPTA) was injected into each knee and MRI signal enhancement was quantified to determine differences in osseous uptake between the two knees. Following MRI assessment, each matched pair was injected with a urethane polymer compound and dissected to correlate vessel disruption with MRI findings. Results: Graft harvest resulted in a mean 31% (range, 7.1-69.5%) decrease in signal enhancement when compared to the matched control. MRI assessment revealed two predominant patterns of vessel entry for the dominant inferior arterial supply. In one pattern, the vessel entered the inferomedial aspect (~7 o’clock) of the distal patellar pole and was disrupted by bone graft harvest in two matched pairs (2/9, 22%). In the second pattern, the predominant vessel entered further medial (~8 o’clock) and was not disrupted in 7 matched pairs. The mean decrease in gadolinium uptake following disruption of the predominant vessel measured 56% (range, 42.6-69.5%) compared to an average decrease of 18% (range, 7.1-29.1%) when the dominant arterial supply to the inferior pole remained intact (p<0.04). The number of mid-patellar vessels that were disrupted along the dorsal surface of the patella was also quantified using MRI and found to correlate with a decrease in signal enhancement (p<0.06) between the experimental and control sides. MRI assessment of mid-patellar vessel disruption was correlated with gross anatomical dissection data to confirm reliability (ICC 0.881, 95% CI: 0.471 to 0.973). Conclusion: Recent anatomic studies have demonstrated that the dominant arterial supply to the patella enters via the inferior pole. We observed two vascular patterns and a medial entry (~8 o’clock) of the predominant vessel precluded disruption by bone graft harvest. Our data suggests that disruption of the dominant arterial supply can result in a significant decrease that exceeds 50% of the total vascular supply to the patella. Further clinical studies are necessary to define the critical threshold of patellar devascularization and its potential role in postoperative morbidity following BPTB harvest.