The anterior cruciate ligament (ACL) is an important structure in maintaining the normal biomechanics of the knee and is the most commonly injured knee ligament. ACL tears are common injuries in all levels of the athletic population¹. Most ACL injuries (~80%) are complete and require surgical repair. A major concern has been raised regarding the difficulty of assessing reconstructed ACL grafts using MRI. Methods based on relative ACL image intensity or proton density is often inclusive. Graft MRI intensity changes might reflect not only graft damage, but also neovascularization and matrix synthesis within the graft.

T₂*-weighted images has been shown to be a good indicator of graft function². However, similar to other tendons and ligaments, the ACL also exhibits a two-compartment T₂* decay profile. The estimated T₂* from single compartment model, or slow decay component from two-compartment model, is sensitive to effect such as edema or other factors unrelated to ACL function^3,4. The fast T₂* (~1 ms) faction originating from water tightly bound with collagen fibers and other tendon structure, is potentially a more sensitive indicator of the intrinsic graft tissue metamorphism. Therefore, an ultrashort echo time (UTE) MRI technique was used in this ACL study to quantify the changes on the T₂* of bound water during ACL graft incorporation.

MRI of the ACL was performed on a Siemens 3T PET/MRI system with a twelve-channel knee coil. A total of eight patients undergoing ACL graft surgery were recruited for this IRB approved study. At this moment, all eight had their ACL imaged at three months post-surgery and two subjects had their ACL imaged at three months and six months post-surgery. 3D double echo UTE imaging was performed with the first echo formed at 100, 270, and 500 µs after non-selective excitation. A spiral out radial projection k-space sampling strategy was used. Other imaging parameters were: FOV of 160 x 160 x 160 mm3; voxel size of 1 x 1 x 2 mm³; TR of 13 ms; flip angle of 15°; total acquisition time of 2.5 mins for each echo.

3D-UTE images were reconstructed using a regridding procedure⁵ adapted for a non-isotropic voxel size. For fast T₂* component calculation an ACL region of interest (ROI) was drawn and the signal was averaged within the ROI for each UTE image. Figure 1 displays the T₂* decay at different TE's, clearly demonstrating the two-compartment feature. However, due to uncompensated eddy current effect, and potential delay of readout gradient and ADC timing, it is un-reliable to combine the data from the first echo (acquired with half readout k-space sampling) and second echo (acquired with full readout k-space sampling) for a complete two-compartment model curve fitting. Therefore, the average signal at each first echo was then fit with a mono-exponential function to calculate the fast T₂* component of the ACL using the equation $$$S(TE)=S(0)\cdot e^{-TE\diagup T2_2^*}$$$.

Since the T₂* of ligaments is dependent on its relative orientation with respect to the scanner B₀ field, only the middle sections of the ACL graft were used for ROI. Across all eight subjects, the mean ACL T₂* of the fast decay component was 1.66 +/- 0.4 ms at their first imaging visit (~3 month after ACL transplant). Figure 2 shows the three UTE images acquired six months post-surgery for two subjects. The average fast T₂* component for the ACL went from 2.55ms at three months post-surgery to 1.81ms at six months post-surgery for the first subject, and 1.39ms to 1.24ms for the second subject. This confirms the hypothesis that tissue metamorphism of ACL graft may reduce T₂* of the ACL bound water. It has been studied that the ACL graft undergoes the process of peripheral synovial vascularization and is on the way to ultrastructural incorporation^6-8.

In this study, the T₂* of the bound water in ACL graft was estimated utilizing a 3D-UTE technique. Based on limited data, the T₂* for the fast component decreases between 3 month to 6 month post-surgery. The results demonstrate the promise of establishing the use of UTE and T₂* mapping of bound water as a biomarker for the ACL graft incorporation process and prediction of patient outcome.