How to optimize the use of MRI in anatomic ACL reconstruction

A 1.5-T magnet is used with an open-bore configuration (Magnetom Espree, Siemens Medical Solutions, Malvern, PA, USA). The ACL is optimally imaged using multiple planes. The different pulse sequences used for evaluating the ACL are outlined in Table 1, including the role of each sequence. The normal ACL on MRI has a hypo-intense appearance on T1- and T2-weighted sequences.

The protocol starts with a scout image in the axial, sagittal and coronal plane. The axial images are true axial and are derived directly from the scout images. The sagittal images are based on anatomic landmarks to individualize the sequence to ensure proper visualization of the ACL and its two bundles for every patient. A plane is prescribed along the lateral femoral epicondyle at the level of the lateral collateral ligament (LCL), which is a constant anatomic landmark with little interpersonal variability. An oblique sagittal sequence is also obtained, which runs parallel to the ACL and optimally visualizes both the AM and PL bundles (Fig. 1). Routine coronal images are then prescribed, followed by a special oblique-coronal sequence (Fig. 1). This plane is in the long axis of the ACL, starting at the intercondylar roof of Blumensaat’s line. The coronal-oblique sequence increases the sensitivity and specificity of diagnosing isolated AM or PL bundle injuries [8, 14]. It also helps to visualize the proximal insertion of the bundles for haemorrhage and rupture. A slice thickness of 2.5 mm can differentiate the two bundles as separate entities or alternative [6]. 3D imaging can also be obtained with the same 1.5-T magnet.

The ACL rupture can be clinically diagnosed most of the time. With the high sensitivity and specificity of physical examination, most ACL ruptures can be diagnosed without the need for further imaging [4]. However, in cases where the physical examination is inconclusive, MRI can provide high degree of diagnostic accuracy [7, 17, 28]. The physical examination may especially be unclear in the case of isolated bundle ruptures [31]. When there is a large Lachman, but only a gliding or no pivot shift, an isolated AM bundle rupture should be expected [19, 20]. Similarly, when there is a large pivot shift, with minimal anterior displacement on Lachman examination, an isolated PL bundle is likely [19, 20]. Confirmation of this diagnosis pre-operative on MRI is essential for planning of treatment (Fig. 2). Partial ACL tears may heal or progress, which can be carefully followed using MRI (Fig. 3). In addition, symptomatic partial ACL tears are suitable to undergo augmentation surgery [5, 21, 24, 26, 30].

The size of the ACL varies greatly amongst individuals [9, 10, 15, 22, 23]. In a clinical study of 137 patients undergoing ACL reconstruction, the tibial ACL insertion site measured 17 mm in length with a standard deviation of 2 mm, while the femoral insertion site measured 16.5 mm in length with a standard deviation of 2 mm [15]. One of the principles of anatomic ACL reconstruction is to restore the native ACL insertion site size and dimensions. This means that a patient’s native ACL insertion site size determines the type of procedure (single- or double-bundle reconstruction), tunnel size and graft size [20]. The tibial insertion site can be measured pre-operatively, which facilitates pre-operative planning (Fig. 4). A single sagittal proton density image best showing the ACL fibre attachment to the tibia is selected. The image is then imported to OSIRIX (Version 3.7.1., Pixmeo Sari, Bernex, Switzerland) for analysis. The most anterior and the most posterior portion of the ACL attachment is marked, and the distance between them is measured. In addition, the length of the native ACL can be measured using the same steps described for the tibial insertion site measurement (Fig. 5). The clearest MRI cut showing both the tibial and femoral insertion sites is chosen, and the distance between the mid-portion of the insertion sites is measured. Determining ACL size may facilitate preoperative planning for an individualized approach to ACL reconstruction in terms of graft options, as well as determining whether the patient will undergo single- or double-bundle surgery. Post-operatively, the reconstructed ACL insertion size can be compared to the native ACL insertion site size to see how much of the ACL has been restored. The same is possible for the intra-articular graft length.

The inclination angle is a simple and easy measurement that can be used for evaluation of graft placement [1, 3, 11]. Measurement of the inclination angle is performed on the sagittal sequence (Fig. 6). The long axis of the tibia is estimated by selecting its most distal portion available in the image. The width of the diaphysis (line a) parallel to the physeal scar is then measured. Proximally in the diaphysis, another width measurement (line b) is taken half of the distance from the most distal width. The mid-points of these lines are connected coursing the long axis of the tibia (line c) and a line perpendicular to the line c, named the tibial horizontal line (THL), is drawn. A parallel line to the THL is created and its end placed at the most anterior portion of the ACL. The angle between this line and the most anterior fibres of the ACL is designated as the ACL inclination angle. Normal ACL inclination angle ranges from 43° to 57° [12]. This method, described by Illingworth et al. [12], provides an intra-observer and interobserver reliability of (ICC) 0.85 (95 % CI .73–.92) and 0.75 (95 % CI .60–.85) for native MRI inclination angle measurements and 0.98 (95 % CI .96–.99) and 0.91 (95 % CI .85–.95) for SB ACL reconstructions, respectively. The sensitivity and specificity, using a ROC curve for anatomic determination, for an inclination angle of 55.0°, were 100 and 87.5 %, respectively. Before the surgery, the native inclination angle can be measured; however, when the ACL is torn, this measurement becomes less accurate. After anatomic single- or double-bundle ACL reconstruction, the inclination angle of the reconstructed ACL can be compared to that of the native ACL. When the ACL is truly reconstructed in an anatomic fashion, the pre- and post-operative inclination angle should be within a few degrees of each other.

This principle can also be applied to revision surgery. The inclination angle can be used to determine whether the previous graft was placed anatomically, by comparing the inclination angle with the ACL inclination of the contralateral limb. If there is a large discrepancy, this means the ACL was most likely placed non-anatomically and use of the existing tunnels during revision surgery should be avoided [30]. After revision surgery, the inclination angle of the graft can be re-measured, to ensure whether satisfactory restoration of the native inclination angle was obtained. The inclination is a simple but reliable method, which is highly correlated with the degree of anatomic tunnel placement [12].

Many graft options exist for ACL reconstruction. Both allograft and autograft material can be used. This decision is based on patient age, activity level, concomitant injuries, previous surgeries, and patient and surgeon preference. Allograft material offers less anaesthesia time, no additional skin incisions, less post-operative pain and free choice of graft diameter and length. Autograft has the benefit of low risk of disease transmission and faster graft healing [16]. The disadvantage of autograft material is that the graft size depends on the size of the donor tissue. However, for bone-patellar tendon-bone (BPTB) and quadriceps tendon autograft, it is possible to evaluate graft size pre-operatively on MRI. On the sagittal proton-density MRI sequence, both tendons can be visualized and their thickness can be assessed (Fig. 7). The slice that demonstrates maximum anterior-to-posterior (AP) thickness of the quadriceps tendon is selected. A line of 15 mm starting from the mid-portion of the superior pole of the patella following quadriceps tendon proximally is drawn. Perpendicular to this line, another line connecting the most anterior to the most posterior portion of the tendon represents the thickness of the quadriceps tendon. The same method is used to determine the patellar tendon thickness using the distal pole of the patella as a starting point.

When the thickness of the tendon is less than 7 mm, there may not be sufficient graft for a double-bundle reconstruction and other graft options including hamstring autograft and allograft should be discussed with the patient.

MRI can also be used for post-operative evaluation after anatomic single- or double-bundle ACL reconstructions. When anatomic double-bundle reconstruction has been performed, the AM and PL bundle can be visualized, evaluated and measured separately. Post-operative MRI evaluation can consist of measuring the size of the restored insertion site and intra-articular graft length.

Recently, there has been an increased use of MRI for evaluation of graft healing. Although much remains unknown about the exact time line of ACL healing on MRI, some global stages can be observed, indicated by variation in signal intensity [2]. The graft remodels between 4 and 8 months after surgery [2]. During this period, revascularization of the neoligament, as well as re-synovialization, takes places. These processes increase the signal intensity [2]. After this initial period, the ACL graft will return to a hypo-intense signal, comparable to the native ACL, which will take approximately 1–2 years [2]. However, this process can differ depending on what graft source is used, for example allograft or autograft material [16].