Summary
Preoperative MRI can be used to predict quadrupled hamstring autograft diameter used in ACL reconstruction.
Abstract
Background
Quadrupled hamstring tendon autograft (HTA) is the most common graft choice for ACL reconstruction (ACLR) worldwide. HTAs with diameter <8 mm used for ACLR have been associated with increased failure rates. Accurate estimates of HTA diameter pre-operatively would help guide optimal ACLR graft selection and patient counseling.
Purpose/Hypothesis:
We aimed to estimate intra-operative HTA diameter based upon pre-operative MRI measurements of potential autograft tendon sources. We hypothesized that pre-operative MRI measurements of potential autograft sources could be used to accurately predict which patients will have grafts > or = 8mm.
Methods
Patients with a history of ACLR with HTA and pre-operative 3-Tesla (3T) MRI performed at our institution were included in this retrospective evaluation. Intra-operative HTA diameter was recorded from each patient’s operative report. Two independent blinded reviewers measured multiple tendon parameters on each patient’s preoperative MRI including semitendinosus (ST) cross-sectional area (CSA), gracilis (G) CSA, patellar tendon (PT) length, PT width, PT thickness, and quadriceps (Q) thickness. CSA was calculated for ST and G using the elliptical region of interest tool on the corresponding proton density–weighted axial image of the knee. The axial slice positioned at the widest portion of the femoral condyle was used for CSA measurements. Bivariate analysis was performed using height, weight, BMI, age, ST CSA, GT CSA, PT length, PT width, PT thickness, and Q thickness as predictors of intra-operative HTA diameter > or = 8mm. The three best predictors were combined into a weighted, additive logistic regression model to determine a threshold score. Reliability statistics between both image readers for all MRI measurements were measured using coefficient of variation.
Results
Application of our inclusion criteria resulted in a total of 53 patients who underwent ACLR with HTA. Height, weight, PT length, PT thickness, ST CSA, and GT CSA were found to be significant predictors for having intra-operative HTA > or = 8mm in diameter. PT length, PT thickness, and GT CSA were the strongest predictors and these parameters were combined into an additive logistic regression model where logit score = [-23.24 + (1.68 * PT length) + (20.104 * PT thickness) + (147.7 * GT CSA)]. If the logit score is >0.237 in our data set, the HT graft diameter is predicted to be > or = 8mm with 87.4% accuracy. Application of a logit function calculation to the logit score can then be used to calculate the probability that the intra-operative HTA is > or = 8mm in diameter. PT length, PT thickness, and GT CSA had coefficient of variance between image readers of 2.0%, 10.9%, and 13.8%, respectively.
Conclusion
Preoperative MRI measurements of PT length, PT thickness and GT CSA can be used to predict quadrupled intra-operative HTA diameter > or = 8mm with a high degree of accuracy and this approach can help guide preoperative planning and ACLR graft selection.