Summary
In young athletes undergoing QT ACLR, the addition of LET does not appear to affect autograft maturation as measured by SI
Abstract
Introduction
LET procedures can reduce strain on the ACL autograft by limiting excessive tibial translation and rotation. Altered loading of the autograft may subsequently impact its maturation, which can be longitudinally evaluated with autograft signal intensity (SI), volume, and knee laxity. However, the effects of LET on both autograft maturation and knee laxity remain unclear.
Methods
Eighteen participants (15.9 ? 1.5 years; 72.2% female) underwent quadriceps tendon (QT) ACLR (Non-LET; N=11) or QT ACLR with LET (LET; n=7) from a prospective ACL registry. MRI was acquired at, 3 months (3M), 6 months (6M), and 12 months (12M) post-operative. Manual segmentation of the autograft was performed to extract graft median signal intensity (SI) and volume (mm3), as well as to inform a future automated pipeline. Specifically, an AI pipeline was developed using a full-resolution 3D nnUNet model with accuracy evaluated using the Sørensen–Dice similarity coefficient to compare to the manual segmentations. Separate 2 (LET Status; QT ACLR, QT ACLR with LET) x 3 (time; 3M, 6M, 12M) mixed ANOVAs with repeated measures and post-hoc t-test with Bonferroni corrections were employed to differentiate post-operative maturation metrics across timepoints and LET status. Additionally, at 12 months, knee laxity was assessed by anterior tibial translation (ATT) in both the reconstructed and contralateral limb using a KT-1000 Arthrometer at a +30 ft-lb force and used to calculate side-to-side ATT difference (SSD ATT). Paired t-tests were used to compare ATT between limbs within each subject and across groups (LET and Non-LET). Separate stepwise linear regressions were preformed to determine whether SI, graft volume, or LET status could predict knee laxity outcomes.
Results
The automated segmentation pipeline achieved dice scores of 83.85 (95% CI: 81.42, 85.84) [Figure 1]. There was only a significant effect for time observed for SI (p<0.001). Post-hoc paired t-tests indicated SI at 12M (Mean= 1.146, SE=0.080) was significantly lower than at 3M (Mean=1.426, SE=0.077, p=0.001) and at 6M (Mean= 1.350, SE=0.071, p=0.005) [Figure 2]. Similarly, there was only a significant effect of time observed for graft volume (p<0.001). Post-hoc paired t-tests indicated graft volume at 12M (Mean =1862.611 mm3, SE 94.287) was significantly lower than at 3M (Mean=2506.541 mm3, SE=105.493, p<0.001), and at 6M (Mean=2346.120, SE=188.812, p=0.022) [Figure 3]. There were no significant differences between LET and non-LET groups in the injured limb ATT, SSD ATT, or differences between Injured limb ATT and uninjured limb ATT (all p’s>0.05) [Figure 4]. Regression analyses indicated that 6M graft volume predicted injured limb ATT with an adjusted r2 of 0.203 [Table 1] and 12M SI predicted SSD ATT with and adjusted r2 of 0.277 [Table 2].
Discussion
Autograft SI and volume decreased over time but was not uniquely affected by the addition of LET. Notably, 12M SI and 6M graft volume were predictors for knee laxity.
CONCLUSION(S): In young athletes undergoing QT ACLR, the addition of LET does not appear to affect autograft maturation as measured by SI. Moreover, both SI and graft volume offer modest predictive value for clinically relevant biomechanical outcome following ACLR.