Predicting Hamstring Graft Diameter Using MRI and Anthropometry: Accuracy With Two Different Graft Preparation Techniques

OBJECTIVES: Graft diameter is one variable that may affect outcome of ACL reconstruction. The ability to predict the size of a graft in a given patient pre-operatively may help guide graft selection and preparation technique. Various papers have correlated anthropometric data and MRI tendon measurements to intraoperative graft diameter, although no papers have investigated these together. The intra-operative diameter of a hamstring autograft will be influenced by graft preparation technique. Our study aimed to investigate the prediction of intraoperative graft diameter of 2 different graft construct techniques (4-strand semitendinosus versus quadrupled semitendinosus) using anthropometry and MRI measurements. METHODS: Retrospective review of two groups of ACL reconstruction using different graft preparation techniques was performed. “Conventional” 4-strand gracilis + semitendinosus with fixed suspension at the femur and screw fixation at the tibia were compared with quadrupled semitendinosus grafts with adjustable suspensory fixation at each end (Graftlink). Cross-sectional areas (XSA) of the semitendinosus and gracilis tendons was measured in the axial slice of a T2 weighted MRI image using a region-of-interest tool. Stepwise linear regression using intraoperative graft diameter as the dependant variable was performed using MRI XSA of the semitendinosus and gracilis tendons, gender and height as predictors. RESULTS: 129 ACL Reconstruction in 127 patients were done in the time period, 89 of which were done conventionally, and 40 which employed the Graftlink construct. The median graft diameter in the Graftlink group (8.5mm IQR8-9) was greater than that of the conventional group (8mm, IQR 7.5-8) (p < 0.001). MRI XSA of semitendinosus and height were statistically significant predictors of diameter in the Graftlink group (R(2) = 51%), whilst MRI XSA of semitendinosus + gracilis and gender were predictors in the conventional group (R(2) = 36%). CONCLUSION: Graftlink ACLRs produced a greater graft diameter than conventional 4-strand semitendinosus and gracilis constructs. Gender and tendon XSA were statistically significant predictors of graft diameter in conventional four-tailed ACLRs, whilst height and tendon XSA predicted Graftlink ACLRs. A greater degree of the graft diameter variance could be predicted pre-operatively in the Graftlink group compared to the conventional construct.