A rationale for predicting graft impingement by the anterior cruciate intercondylar roof A magnetic resonance imaging study

This study was designed to analyze how anterior tibial tunnel placement can result in graft impingement by the intercondylar roof. The relationship of the ACL to the intercondylar roof was studied using magnetic resonance scans. An attempt was made to predict the amount of bone that may need to be removed from the intercondylar roof to prevent impingement on a 10 mm thick ACL graft. Magnetic resonance scans of 19 normal ACLs were analyzed. The amount of bone removal required to correct roof impingement was determined for a graft placed either eccentrically or centrally within the ACL insertion, and within the bulk of the normal ACL fibers. An eccentric tibial tunnel placement required approximately 5 to 6 mm and a central placement required 2 to 3 mm of bone removal from the intercondylar roof to prevent impingement. Placing the graft within the bulk of the ACL fibers, just 3 mm posterior to the center of the ACL insertion, required little bone resection to prevent impingement. To prevent ACL graft impingement, roofplasties need to be performed in both acute and chronic ACL reconstructions if the presently accepted locations for positioning the tibial tunnel are used. A more anteriorly placed tibial tunnel requires more bone removal to prevent roof impingement than a more posteriorly positioned tibial tunnel. The assessment and treatment of intercondylar notch impingement of an ACL graft is a complex problem. Bony impingement can occur from encroachment of the graft by the lateral condyle when the notch is stenosed or when the tibial tunnel is positioned too far laterally? This form of impingement occurs in the coronal plane and results when the dimensions of the graft extend beyond the boundaries provided by the width of the notch. Bone can be removed from the medial wall of the lateral femoral condyle to correct this form of impingement. Bone removal from this location can be more specifically termed a “wallplasty,” instead of using the generalized term “notchplasty.” Guidelines for performing a wallplasty have been established for the chronic ACL deficient knee. The indications for bone removal in the acutely injured knee have not yet been defined. Odensten and Gillquist recommend widening the aperture of the intercondylar notch in the coronal plane to at least 21 mm. Intraoperative visual inspection of the proximity of the graft to the lateral femoral condyle can also be used to determine when a lateral wallplasty is needed and whether it has been completely performed. Graft impingement can also be related to variations in the placement of the tibial tunnel in the sagittal plane. Impingement in the sagittal plane occurs when the intercondylar roof impacts on the ACL graft before the knee reaches terminal extension. 25 Anatomical and roentgenographic studies have confirmed that the normal ACL abuts the intercondylar roof with the knee in full extension. There is normally a broad, anterior flare to the ACL at its tibial insertion that extends anterior to the slope of the intercondylar roof.20 A tubular graft centered anteriorly within this anterior flare of the ACL insertion can impact against the intercondylar roof before the knee reaches terminal extension. This form of impingement in the sagittal plane is “roof impingement” and is anatomically distinct from the impingement in the coronal † Address correspondence and reprint requests to: Stephen M. Howell, MD, 7601 Timberlake Way, Suite 103, Sacramento, CA 95823, plane produced by the medial wall of the lateral femoral condyle. The indications for performing a “roofplasty” in the acute and chronic ACL deficient knee have not been agreed upon. This may be due, in part to the lack of a consensus on the optimal location for placing the tibial tunnel. Clancy et al. state that an eccentric placement (5 mm anterior and medial to the center of the ACL insertion) is preferable. In contrast, Odensten and Gillquist recommend placing the tibial tunnel in the center of the ACL insertion. Further variation in tunnel placement occurs because precise placement of the tibial tunnel is difficult to achieve. The lack of a consensus on a recommended tibial tunnel location, and imprecision in tunnel placement, indicate that there is likely to be significant variation in tibial tunnel placement between surgeons and within successive reconstructions performed by the same surgeon. Placement variation of the tibial tunnel in the sagittal plane will produce different degrees of roof impingement. Detection of roof impingement can be difficult as it occurs only when the knee approaches the last few degrees of terminal extension. The trochlea of the femur closes on the tibial plateau as the knee extends and obstructs the view of the relationship of the intercondylar roof to the graft. Knee surgeons continue to report complications of ACL reconstructions such as pain, limited knee extension, synovitis, graft abrasion, and late graft rupture, which can be attributed in some degree to graft impingement. Additional studies are needed to analyze the different factors that contribute to graft impingement. This study was designed to determine how different sagittal positions of the tibial tunnel may affect the severity of intercondylar roof impingement. Magnetic resonance (MR) scanning was used to provide detailed, in vivo visualization of the ACL including its origin, its intraarticular relationship to the intercondylar notch, and the sagittal limits of its tibial insertion. The objectives of this study were to estimate the amount of bone removal from the intercondylar roof required to prevent roof impingement if a graft is placed either 1) eccentrically or 2) centrally within the ACL insertion. Thirdly, and of equal interest, was the tibial location where a graft could be positioned to repIace the bulk of the ACL fibers and yet still be free from impingement without bone removal. These three objectives were studied by analyzing the anatomical relationship of the ACL and intercondylar roof using MR scans of the normal knee. MATERIALS AND METHODS Sagittal MR scans from 42 consecutive patients with an intact ACL determined by history and physical examination were retrospectively reviewed. Twenty-three patients were excluded because the dimensions of the normal ACL were not completely displayed on one MR image. (These patients had the ACL projected on two or more slices.) The remaining 19 patients, with either normal knees or stable knees with isolated meniscal lesions, comprised the study group. The knees with meniscal lesions displayed no evidence of arthritic changes on AP, lateral, and notch radiographs. There were 15 males and 4 females; the average age was 33 ± 11 years (range, 16 to 48 years). Imaging was performed using a Philips 1.5 Tesla superconducting magnet with a dedicated surface receiver coil. Images were obtained using contiguous 2.5 mm thick sagittal sections (0.625 mm pixels) centered about the ACL. The fully extended knee was externally rotated 10° to 15° in an attempt to optimally align the ACL in the sagittal plane (although this maneuver resulted in a true sagittal alignment in only 19 of 42 patients). Extension placed the ACL against the intercondylar roof. Image acquisition was performed with the standard spin-echo technique using a repetition time (TR) of 1200 msec and an echo time (TE) of 40 msec. Encoding and reconstruction was performed with the standard two-dimensional Fourier transformation technique, using 256 phase encoding steps and two excitations (10 minutes acquisition time). The sagittal image that best displayed the normal ACL was magnified to 140% of acutal size (magnification factor of 1.4). Measurements were performed on the single image that depicted the complete sagittal dimensions of the ACL from origin to insertion. Measurements were made of the film rather than directly from the monitor in order to determine the reproducibility of obtaining these measurements in a setting easily accessible by an orthopaedic sur-geon. Anatomical relationships between the tibial plateau, the ACL, and the intercondylar roof were obtained by using the following measurement technique to analyze the MR scans (Fig. 1A). The midsagittal depth of the tibial plateau was measured by drawing a line perpendicular to the long axis of the tibia centered at the level of the anterior edge of the ACL insertion. A perpendicular line was placed along the anterior tibial cortex to define the anterior limit of the tibial plateau. Another perpendicular line was drawn posteriorly against the tibial cortex that overlies the tubercle near the PCL insertion. The sagittal depth of the tibia was measured between these limits and converted to actual size using the magnification factor. The relationship of the ACL to the tibial plateau and the intercondylar roof was studied by selecting several landmarks (Fig. 1B). The anterior-most portion of the ACL insertion was marked at the joint line (A). A line was drawn along the anterior edge of the ACL paralleling the relatively straight proximal one half of the ligament. A second line, parallel to the anterior line, was drawn overlying the posterior edge of the ACL. The region between these two lines defined the course of the bulk of the ACL ligament fibers. The width of the bulk of the ACL was measured between these two lines at the midpoint between the origin and insertion of the ACL (W) and converted to actual size using the magnification factor (x1.4). The intersection of the posterior line with the joint line marked the posterior limit of the ACL insertion (P). The relationships of the central and eccentric placements Vol. 19, No. 3, 1991 ACL Graft Impingement: MRI Study 277