- Email: [email protected]
Comparison of Anterior Cruciate Ligament Tunnel Position and Graft Obliquity With Transtibial and Anteromedial Portal Femoral Tunnel Reaming Techniques Using High-Resolution Magnetic Resonance Imaging
Comparison of Anterior Cruciate Ligament Tunnel Position and Graft Obliquity With Transtibial and Anteromedial Portal Femoral Tunnel Reaming Techniques Using High-Resolution Magnetic Resonance Imaging Andrea L. Bowers, M.D., Asheesh Bedi, M.D., Joseph D. Lipman, M.S., Hollis G. Potter, M.D., Scott A. Rodeo, M.D., Andrew D. Pearle, M.D., Russell F. Warren, M.D., and David W. Altchek, M.D.
Purpose: Using 3-dimensional high-resolution magnetic resonance imaging (MRI), we sought to compare femoral and tibial tunnel position and resultant graft obliquity with single-bundle anterior cruciate ligament (ACL) reconstruction using transtibial (TT) or anteromedial (AM) portal femoral tunnel reaming techniques. Methods: Thirty patients were prospectively enrolled after primary, autogenous bone–patellar tendon– bone ACL reconstruction by 2 groups of high-volume, fellowship-trained sports medicine surgeons. With the TT technique, an external starting point was used to maximize graft obliquity and femoral footprint capture. By use of high-resolution MRI and imaging analysis software, bilateral 3-dimensional knee models were created, mirrored, and superimposed. Differences between centroids for each femoral and tibial insertion, as well as corresponding ACL/graft obliquity, were evaluated with paired t tests and 2-sided Mann-Whitney nonparametric tests, with P ⬍ .05 defined as significant. Results: No significant differences were observed between groups in position of reconstructed femoral footprints. However, on the tibial side, AM centroids averaged 0.8 ⫾ 1.9 mm anterior to native ACL centroids, whereas the TT group centered 5.23 ⫾ 2.4 mm posterior to native ACL centroids (P ⬍ .001). Sagittal obliquity was closely restored with the AM technique (mean, 52.2° v 53.5° for native ACL) but was significantly more vertical (mean, 66.9°) (P ⫽ .0001) for the TT group. Conclusions: In this clinical series, AM portal femoral tunnel reaming more accurately restored native ACL anatomy than the TT technique. Although both techniques can capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. When a tibial tunnel is drilled without the need to accommodate subsequent femoral tunnel reaming, more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. Level of Evidence: Level III, retrospective comparative study.
E
ndoscopic anterior cruciate ligament (ACL) reconstruction with transtibial (TT) preparation of a femoral socket has been performed for nearly 3 decades with good clinical success. With improved un-
derstanding of the complex anatomy and function of the native ligament, however, modifications to the TT technique have been suggested to improve femoral tunnel position and restore the native ligament foot-
From the Sports Medicine and Shoulder Service, Hospital for Special Surgery, New York, New York; and MedSport, Department of Orthopaedic Surgery, University of Michigan (A.B.), Ann Arbor, Michigan, U.S.A. Supported by the Institute for Sports Medicine Research. The authors report no conflict of interest. Received March 21, 2011; accepted July 6, 2011. Address correspondence to David W. Altchek, M.D., 535 E 70th St, New York, NY 10021, U.S.A. E-mail: [email protected] © 2011 by the Arthroscopy Association of North America 0749-8063/11180/$36.00 doi:10.1016/j.arthro.2011.07.007
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 27, No 11 (November), 2011: pp 1511-1522
1511
1512
A. L. BOWERS ET AL.
prints and graft obliquity.1,2 Nonetheless, questions have been raised regarding the ability of the TT technique to accurately and safely re-create the correct femoral and tibial insertion sites.3-6 In the past decade, many investigators have advocated preparation of the femoral tunnel independently through an anteromedial (AM) portal or 2-incision technique, thereby avoiding the constraints of tibial tunnel drilling.7-13 Preliminary navigation studies in cadavers have shown improved tunnel position and graft kinematics with techniques that use independent femoral tunnel drilling with ACL reconstruction.6,7,14-20 Clinical comparisons of the 2 techniques, however, are limited.21 A previous magnetic resonance imaging (MRI) study of graft obliquity by Hantes et al.22 has been limited to use of isolated 2-dimensional slices of tunnels in an attempt to characterize a 3-dimensional (3D) structure. Recently established was a new, validated method for tunnel evaluation using 3D MRI models that compare a mirror-image overlay of the center of the femoral tunnel aperture with the center of the native ACL attachment on the contralateral, native knee.14 Such models were generated by Abebe et al.14 to show that independent drilling with an outside-in guide pin and use of a retrograde drill (Arthrex, Naples, FL) resulted in a femoral tunnel positioned significantly closer to the center of the native ACL whereas the TT technique resulted in femoral tunnels often centered outside of the native footprint. We favor the use of this 3D technique because it allows for direct comparison with the patient’s own native ACL anatomy and determination of the exact fiber insertion rather than simply the tunnel position. This technique minimizes errors due to signal averaging and through-plane resolution when evaluating ACL anatomy. To our knowledge, however, there are no clinical reports in the published literature using 3D MRI to directly compare TT and AM portal drilling techniques. The specific aim was to compare femoral and tibial tunnel position and resultant graft obliquity achieved with single-bundle ACL reconstruction using a TT or AM portal reaming technique. We evaluated 3 main hypotheses: 1. Femoral tunnel position would be significantly improved and closer to the native femoral ACL attachment with an AM portal compared with a TT reaming technique. 2. Tibial tunnel position would be significantly posterior relative to the native tibial ACL footprint with the TT technique compared with the AM portal technique.
3. The resultant ACL graft would be more vertical in both the sagittal and coronal planes with the TT technique compared with the AM portal technique. METHODS Patient Selection and Study Design This study was approved by our Institutional Review Board (Hospital for Special Surgery, No. 29116). Inclusion criteria comprised patients aged 16 to 40 years who underwent elective, primary ACL reconstruction with autograft bone–patellar tendon– bone. Exclusion criteria were previous surgery on either the ipsilateral or contralateral knee, multiligamentous reconstruction, concomitant bony procedures that might distort the bony anatomy (osteotomy, meniscal transplant, osteoarticular allograft or autograft), patients unwilling or unable to undergo MRI, or patients with contraindications to MRI (e.g., pacemaker or certain implants). Subjects were identified prospectively, and data were gathered in a cross-sectional manner during the postoperative period. Fifteen subjects were enrolled for each technique (TT and AM). Operative notes for each were collected and scrutinized to (1) confirm the method used for femoral tunnel preparation (TT or AM) and (2) confirm that no concomitant ligamentous or bony procedure had been performed. A diagram of enrollment and protocol is depicted in Fig 1. ACL reconstructions were performed by 1 of 8 highvolume, fellowship-trained sports medicine surgeons who attempted to optimize femoral tunnel position and
FIGURE 1. Subject selection criteria, enrollment, and protocol. (B-PT-B, bone–patellar tendon– bone).
ACL TUNNEL POSITION AND GRAFT OBLIQUITY re-create native graft obliquity. TT reconstructions were performed with standard rigid reamers for each tunnel. The tibial tunnel was positioned within a portion of the tibial footprint, centered between the tibial eminences, while protecting the anterior horn of the lateral meniscus. Coronal angulation of the guide and resultant external tibial positioning were ultimately selected to allow anticipated closest access to the anatomic center of the femoral footprint through appropriate rotation of an endofemoral guide. Notchplasty was performed as needed to improve visualization, and while viewing with a 30° arthroscope, the surgeon passed a rigid guide pin and reamer through the center of the native femoral ACL footprint for preparation of the femoral tunnel. Our AM technique has been described previously.23 In brief, AM reconstructions were performed with positioning of a tibial guide and resultant tunnel within the native ACL footprint centered both anteroposteriorly and mediolaterally, without reference to surrounding anatomic landmarks. Femoral tunnel preparation was performed through an AM portal positioned immediately above the medial meniscus while allowing direct access to the lateral notch. A gentle notchplasty was performed to improve visualization. While viewing with a 70° arthroscope, the surgeon marked the center of the anatomic footprint with an awl, and with the knee held in approximately 115° of flexion, a flexible guidewire was passed from this point in a superolateral direction to maximize tunnel length. A flexible reaming system (Smith & Nephew, Andover, MA) was then used for tunnel preparation. Three-Dimensional MRI Reconstructions Magnetic resonance images were obtained between 5 and 25 weeks postoperatively. All scans were performed on a clinical 3-T scanner (GE Healthcare, Waukesha, WI) by use of an 8-channel phased-array transmit/receive knee coil. A magnetic resonance image of the contralateral knee was also acquired to serve as a control data set, and there was no history of ACL injury on the unaffected side. Each knee was placed in standardized slight flexion and external rotation during scanning. Morphologic changes of the joint were assessed from 2-dimensional fast-spin echo images acquired along 3 anatomic planes (echo time, 25 to 30 milliseconds; repetition time, 4,000 to 6,000 milliseconds; echo train length, 8 to 16; bandwidth, ⫾62.5 kHz over entire frequency range; acquisition matrix, 512 ⫻ 416 to 512 ⫻ 481; number of excitations, 1 to 2; field of view, 15 to 16 cm; slice thickness, 3.5 mm with no gap).
1513
In addition, a 3D fast-spin echo technique was used to acquire a data set with near-isotropic voxels (echo time, 36 milliseconds; repetition time, 2,500 milliseconds; echo train length, 64; bandwidth, ⫾41.67 kHz over entire frequency range; acquisition matrix, 256 ⫻ 256; number of excitations, 0.5; field of view, 18 cm; slice thickness, 0.6 mm; scan time, 7 minutes). Tissue contrast was provided to ensure differential contrast between ACL graft, bone, fluid, and cartilage. With the operator blinded to the technique used, the magnetic resonance images were segmented and 3D models were then created by use of Mimics 13.1 software (Materialise, Leuven, Belgium). The 3D models were exported to Geomagic Studio 11 (Geomagic, Research Triangle Park, NC). A mirror image of the left side was created. Multiple, corresponding anatomic points between the 2 femurs were registered, and the mirrored left side was rotated and translated for a best-fit overlay onto the right femur. The rotational and translational transformation was then applied for the left-side ACL or graft femoral footprint. The same technique of registration, overlay, and footprint transformation was then applied to the tibial side. Femoral and tibial sides were registered individually to avoid any error that could be generated if there were any differences between sides in terms of sagittal attitude (slightly flexed posture on one side v full extension on the other side at the time the MRI sequence was acquired). Determination of Centroids and Graft Obliquity The centroids of the femoral and tibial insertions of both the ACL and graft were then determined. To describe the positions of the reconstructed centroids relative to native centroids, and to compare between subjects, tunnel positioning was assessed relative to reproducible anatomic landmarks on both sides of the joint. On the femoral side, best-fit spheres were identified for the contour of each of the medial and lateral distal femoral condyles (Fig 2), and the centers of these spheres were used to determine the geometric center axis, as described and validated previously by Asano et al.24 This provided a medial-lateral reference axis. The anterior-posterior axis was calculated perpendicular to the medial-lateral axis. The proximaldistal axis was determined by calculating 2 centroids of cross sections of the periosteal surface of the femur 30 mm apart up the femoral shaft. This established a coordinate system by which the distance between the centroids of the reconstructed graft and native ACL could be described in both the anterior-posterior and
1514
A. L. BOWERS ET AL.
FIGURE 2. Best-fit spheres are identified for the contour of each of the medial and lateral distal femoral condyles. The medial-lateral reference axis was determined by the geographic centers of the spheres, as described and validated previously by Asano et al.24 The anterior-posterior axis was calculated perpendicular to the medial-lateral axis.
proximal-distal directions, measured in millimeters (Fig 3). On the tibial side, the 3D model was rotated so that the articular surface could be viewed en face. The medialand lateral-most points were registered. Next, a best-fit plane was created to match the surface of the plateau medially and laterally. The anterior-posterior axis was determined perpendicular to the medial-lateral axis in this plane. The distance of the tibial centroid of the reconstructed graft could then be described in both anterior-posterior and medial-lateral directions relative to the native tibial footprint, again measured in millimeters (Fig 3).
FIGURE 3. Sample reconstructions for the femur (left) and tibia (right). Distances were measured (in millimeters) between the native and reconstructed centroids in the anterior-posterior and proximal-distal directions for the femur and in the medial-lateral and anterior-posterior directions for the tibia. The native footprints are depicted in green and the reconstructed footprints in red.
Next, the obliquities of the native ACL and reconstructed grafts were measured in the sagittal and coronal planes by use of Pro/Engineer Wildfire 4.0 (Parametric Technology, Needham, MA). These were determined by the angle subtended by a line extending from the centers of the femoral and tibial footprints and the best-fit plane of the tibial surface (Fig 4). Footprint Overlap On the tibial side, the extent of overlap of the surface area for the graft insertion and the native footprint was calculated. By use of ImageJ 1.40g
ACL TUNNEL POSITION AND GRAFT OBLIQUITY
1515
FIGURE 4. Sample measurements for coronal (left) and sagittal (right) graft obliquities. The native footprint and ACL obliquity are depicted in green and the reconstructed graft in red.
(National Institutes of Health, Bethesda, MD), the percent overlap between the native and reconstructed sides was then calculated to determine what portion of the graft insertion fell within the native footprint. Statistical Analysis A sample size of 15 was determined for each group (AM and TT) to achieve 80% power to detect a difference of 3 mm from the center of the native ACL between the 2 approaches, with a conservatively estimated SD of ⫾3 mm and a significance level of .050, as well as adjustment for a 2-sided Mann-Whitney nonparametric test. Graft placement relative to the native ACL was then compared between techniques by use of paired t tests. Native ACL obliquity in the sagittal and coronal planes between the TT and AM groups was compared by use of unpaired t tests to confirm that there were no differences between controls for each group. A graft obliquity difference of 10° from the native ACL in the sagittal plane was considered clinically significant. Notably, 4 scans were segmented and processed in duplicate to verify accuracy; each yielded measurements reproducible within 0.8 mm and 1.0°. Because of the time-intensive nature of the segmentation and processing (approximately 8 hours per subject), the remaining scans were not analyzed in duplicate.
RESULTS We contacted 35 consecutive subjects meeting the inclusion/exclusion criteria for participation; 5 (2 AM and 3 TT) declined participation, citing the time required to sit through bilateral knee MRI. Thus 30 subjects were consecutively enrolled, 15 in the AM group and 15 in the TT group. There were 16 male and 14 female patients. Of the reconstructed knees, 19 were right sided and 11 left sided. The mean age at reconstruction was 24 years (range, 16 to 38 years). The mean length of time from reconstruction to postoperative MRI was 12 weeks (range, 5 to 25 weeks). Data for each subject are depicted in Table 1. Centroid Position On the femoral side, centroids from the AM portal group averaged 1.97 mm anterior (range, 1.60 mm posterior to 4.52 mm anterior), 1.91 mm distal (range, 2.08 mm proximal to 4.20 mm distal), and 2.89 mm (range, 0.50 to 5.90 mm) in greatest distance (hypotenuse) from the matched-pair, native ACL femoral centroids. Centroids prepared with the TT technique averaged 3.02 mm anterior (range, 2.60 mm posterior to 9.91 mm anterior), 2.46 mm distal (range, 2.77 mm proximal to 6.53 mm distal), and 2.15 mm (range,
1516
A. L. BOWERS ET AL. TABLE 1.
Data Collected for 30 Study Participants Difference in Graft Obliquity (°) Relative to Paired, Native ACL
Footprint Centroid Position Femoral
Subject No. Anteromedial group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 TT group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Tibial Maximum Distance (mm)
Direction Relative to Paired, Native ACL
Sagittal
Coronal
Tibial Footprint: Graft Overlap With Native Footprint (%)
Anterodistal Anterodistal Anterodistal Posterodistal Posterodistal Posteroproximal Posteroproximal Posteroproximal Anterodistal Anterodistal Anterodistal Posterodistal Anteroproximal Anterodistal Anterodistal
5.05 4.22 0.76 2.59 2.08 3.39 6.87 3.34 2.42 1.79 2.20 5.08 0.42 1.20 2.78
Anteromedial Anteromedial Posteromedial Anteromedial Posteromedial Posteromedial Anteromedial Anteromedial Anterolateral Posteromedial Posteromedial Posteromedial Anteromedial Anteromedial Anteromedial
⫺0.2 ⫺0.1 ⫺5 7.8 4.7 ⫺5.9 7.7 2.1 0.1 0.5 ⫺2.7 ⫺0.1 ⫺0.5 0.9 ⫺2.3
13.3 20 ⫺0.7 15.7 12.9 10.4 11.9 3 ⫺0.5 6.6 11 14.2 0.3 2.7 12
72.1 56.3 65.3 83.6 63.7 67.5 48.3 42.4 79.5 52.9 55.6 63.4 77.3 89.8 76.6
Anterodistal Posteroproximal Anterodistal Anteroproximal Anteroproximal Anteroproximal Anterodistal Anterodistal Posterodistal Anterodistal Anterodistal Anterodistal Anterodistal Anterodistal Anterodistal
5.67 6.27 11.00 9.68 2.61 4.29 5.34 2.04 7.09 7.56 5.21 3.06 5.88 6.70 4.47
Posteromedial Posteromedial Posterolateral Posteromedial Posteromedial Posterolateral Posteromedial Anteromedial Posteromedial Anteromedial Posteromedial Posteromedial Posteromedial Posteromedial Posterolateral
⫺11.4 ⫺11.2 ⫺27.5 ⫺17.3 ⫺15.8 ⫺13.1 ⫺12.9 ⫺3 ⫺8.1 8.9 ⫺25.2 ⫺3.4 ⫺16.7 ⫺23.1 ⫺7.5
13.7 8.8 ⫺0.7 13 3.7 ⫺4.1 6.7 ⫺2.3 7.5 13.4 6.3 4.7 ⫺10.8 2.9 ⫺2.4
29.4 36 5.6 6 44.4 57.1 55.6 74.5 39.1 29.1 42.2 43.1 39.2 19 48.9
Maximum Distance (mm)
Direction Relative to Paired, Native ACL
0.50 5.90 5.62 3.69 2.23 1.70 1.06 2.08 5.00 1.14 3.71 2.52 1.48 1.63 5.11 1.65 1.46 10.72 2.19 4.56 2.60 3.88 3.37 5.18 0.83 11.19 4.23 4.08 3.52 2.80
0.83 to 11.19 mm) in greatest distance from the paired, native ACL femoral centroids. There were no statistically significant differences between the 2 groups with respect to distance from the native centroid in the anterior-posterior (P ⫽ .206) or proximaldistal (P ⫽ .344) direction or for the hypotenuse (P ⫽ .176) (Table 2). Femoral data points are depicted in Fig 5. On the tibial side, centroids from the AM portal group averaged 1.31 mm anterior (range, 1.52 mm posterior to 6.09 mm anterior), 2.41 mm medial (range, 5.01 mm medial to 2.07 mm lateral), and 2.83
mm (range, 0.42 to 6.87 mm) in greatest distance (hypotenuse) from the paired, native ACL tibial centroids. TT centroids averaged 5.23 mm posterior (range, 11.00 mm posterior to 6.54 mm anterior), 1.91 mm medial (range, 5.45 mm medial to 1.27 mm lateral), and 5.75 mm (range, 2.04 to 11.00 mm) in greatest distance from paired, native ACL tibial centroids. These findings were significantly different for the anterior-posterior plane (P ⬍ .001) and greatest distance (P ⫽ .001) but not for the medial-lateral plane (P ⫽ .344) (Table 2). Tibial data points are depicted in Fig 6.
ACL TUNNEL POSITION AND GRAFT OBLIQUITY
1517
TABLE 2. Comparison of Tunnel Positions Between AM Portal and TT Femoral Tunnel Reaming Techniques Insertion Direction Femur Anterior/posterior Proximal/distal Hypotenuse Tibia Anterior/posterior Medial/lateral Hypotenuse
Technique
Mean Distance (SD) (mm)
AM TT AM TT AM TT
1.97 (1.52) 3.05 (2.87) 1.91 (1.35) 2.46 (1.73) 2.89 (1.80) 4.15 (3.02)
.206
AM TT AM TT AM TT
1.31 (1.54) 5.23 (2.40) 2.41 (1.40) 1.91 (1.44) 2.83 (1.84) 5.75 (2.36)
⬍.001
P Value
.344 .176
.344 .001
Graft Obliquity In the sagittal plane, native ACL obliquity averaged 53.5°, and no significant difference was observed between the native ACL obliquity in the TT group (52.6°) and AM group (54.4°) (P ⫽ .430). Mean graft obliquity in the AM portal group was 52.2° (range, 44.3° to 60.3°) and was not significantly different from that of the paired, native ACL (P ⫽ .653). TT-reconstructed graft obliquity, however, averaged 66.9° (range, 45.5° to 78.4°) and was
FIGURE 6. Tibial position of reconstructed footprint centroids relative to native footprint centroids. A representative composite of native tibial footprints is marked in green, with its centroid marked in gray. Centroids of grafts reconstructed with the TT technique are marked in red, and those with the AM portal technique are marked in blue. The TT technique resulted in a position significantly posterior to centroids for both native ACLs and the AM portal technique.
significantly less oblique than that of the native ACL (P ⫽ .0001) (Table 3). In the coronal plane, native ACL obliquity averaged 76.1°, and no difference was observed between native ACL obliquity in the TT group (77.1°) and AM group (75.0°) (P ⫽ .326). AM-reconstructed graft coronal obliquity averaged 66.1° (range, 58.2° to 77.5°) and was significantly more oblique than that of the paired, native ACL (P ⬍ .0001). TT-reconstructed graft obliquity averaged 73.0° (range, 64.6° to 79.1°) and
TABLE 3. Comparison of Graft Obliquities Between AM Portal and TT Femoral Tunnel Reaming Techniques Obliquity Sagittal Native ACL AM TT
FIGURE 5. Femoral position of reconstructed footprint centroids relative to native footprint centroids. A representative composite of native femoral footprints is marked in green, with its centroid marked in gray. Centroids of grafts reconstructed with the TT technique are marked in red, and those with the AM portal technique are marked in blue. No significant differences were seen among centroid positions between these techniques.
Coronal Native ACL AM TT
Technique
Mean Angle (SD) (°)
AM TT Reconstructed Native Reconstructed Native
52.6 (4.33) 54.4 (7.15) 52.2 (5.07) 52.6 (4.33) 66.9 (9.11) 54.4 (7.15)
.430
AM TT Reconstructed Native Reconstructed Native
75.0 (5.67) 77.1 (5.62) 66.1 (4.69) 75.0 (5.67) 73.0 (4.50) 77.1 (5.62)
.326
P Value
.653 .0001
⬍.0001 .045
1518
A. L. BOWERS ET AL.
FIGURE 7. Maximum, mean, and minimum tibial footprint fill for AM portal and TT techniques. Native footprints are depicted in green, and reconstructed footprints are shown in red. Tibial graft insertion prepared with the AM technique had a mean of 66.3% overlap with the native ACL tibial footprint. For the TT technique, the overlap was 38.0% (P ⬍ .0001).
also was somewhat greater than that of paired, native ACLs (P ⫽ .045) (Table 3). Tibial Footprint Fill With respect to footprint fill, tibial graft insertion prepared with the AM technique had a 66.3% overlap with the native ACL. The portion that did not overlap fell predominantly medial to the native insertion. For the TT technique, the overlap was 38.0% (P ⬍ .0001); the remaining 62% fell predominantly posterior and slightly medial compared with the native footprint (Fig 7). DISCUSSION Independent drilling of the femoral tunnel through an AM portal compared with TT reaming has been theorized to better restore native anatomy in ACL reconstruction. We hypothesized that reconstructions performed with AM portal reaming would yield more accurate femoral position, tibial position, and graft obliquity. In this cross-sectional MRI study, we found that both techniques performed equally well with respect to femoral tunnel position. However, the AM portal technique resulted in more anatomic tibial tunnel position and sagittal graft obliquity. True “anatomic” ACL reconstruction remains elusive because of the complex native anatomy with differing fiber lengths. Nonetheless, ACL reconstruction with attachment points that more closely matched native ligament attachment location has been repeatedly shown to better reproduce knee kinematics in both cadaveric and in vivo models6,13,15,25-27 and is theorized to improve clinical outcomes. The surgical technique by which to prepare tibial and femoral sockets while respecting the native ACL anatomy, however, is controversial. Whereas some authors have recommended preparation of the femoral tunnel using a TT technique, others have advocated for indepen-
dent femoral tunnel drilling through an AM arthroscopic portal. Endoscopic reconstruction techniques have evolved in an effort to increase graft obliquity, because the relatively vertical graft prepared with traditional TT techniques succeeded in limiting anterior translation but not necessarily the rotatory instability seen in ACL insufficiency.25,28 Traditional TT techniques as described by Morgan et al.29 suggested use of internal landmarks to select a relatively posterior position on the tibia, resulting in femoral tunnel position in the proximal portion of the femoral footprint. Increasing the coronal obliquity2 or using external landmarks to select a more medial starting point1 in preparation of the tibial tunnel has been suggested to improve subsequent positioning of the femoral tunnel closer to the native anatomic ACL insertion on the femur. In navigated cadaveric specimens, femoral tunnels near the center of the femoral insertion have been shown to be achievable through a TT technique; however, these required meticulous tunnel placement “at the limits of practical,” aperture beveling, and the acceptance of some graft-tunnel mismatch for autograft bone– patellar tendon– bone grafts.30 Despite these modifications, studies have repeatedly shown that tunnel positioning remains compromised with TT ACL reconstruction techniques.3,4,31-33 Brophy et al.15 and Pearle et al.20 reported that the traditional, arthroscopic TT technique predisposed to a “mismatch” graft position from the posterolateral (PL) tibial footprint to the AM femoral footprint. Heming et al.5 found that a TT technique could produce tunnels centered in both the tibial and femoral footprints only if a starting point prohibitively close to the joint line with a correspondingly short tibial tunnel is used. Meanwhile, direct comparisons of femoral tunnel position and graft obliquity after TT or AM portal ACL reconstruction are growing in the literature. In both cadaveric6,17 and in vivo3,22 studies, femoral tun-
ACL TUNNEL POSITION AND GRAFT OBLIQUITY TABLE 4.
1519
Relevant Historical Analyses of TT Reconstructions Compared With Native ACL and AM Reconstructions
Study
Analysis
Arnold et al.31
Cadaveric arthroscopy and dissection
Giron et al.4
Cadaveric dissection and plain radiographs
Dargel et al.3
Clinical reconstruction with plain radiographs
Ahn et al.32
Clinical reconstruction with 2-dimensional MRI
Stanford et al.33 Gavriilidis et al.17
Computer navigation of cadavers v clinical reconstructions Cadaveric arthroscopy and dissection
Hantes et al.22
Clinical reconstruction with 2-dimensional MRI
Steiner et al.6
Cadaveric navigation and kinematics
nel position and graft obliquity have been shown to be more closely replicated with AM compared with TT techniques. Relevant historical studies are summarized in brief in Table 4. In our series TT reconstructions were performed by high-volume (50 to ⬎150 ACL reconstructions per year), fellowship-trained sports medicine surgeons. The goal was anatomic restoration of the native ACL footprint and ligament obliquity in all cases. We found that the femoral tunnel achieved with a TT technique was similar in position to that created with AM techniques. Our results were similar to those seen in a select few publications.34,35 Giron et al.34 reported that both TT and AM techniques were able to re-create a reference hole made just deep to the native ACL AM bundle insertion in cadaveric knees. Albuquerque et al.35 compared 10 cadaveric reconstructions performed through a TT technique with 10 performed using AM drilling. In 3 outcomes measured (posterior wall thickness, tunnel position at notch, and tunnel inclination in relation to femoral axis), no differences were observed between the 2 techniques. Contrary to our first hypothesis, it appears that improved femoral tunnel position is possible with a TT technique. However, in an attempt to produce an anatomic reconstruction, the tibial tunnel position is also of great importance, because we recognize that the positions of both tunnels ultimately determine graft obliquity with a TT technique. Posterior placement of the tibial tunnel relative to the native tibial footprint has
Pertinent Findings Using a correctly placed tibial tunnel, an anatomic attachment on the femur could not be achieved through a TT technique Impossible to restore anatomic femoral ACL origin with TT drilling technique TT drilling resulted in femoral tunnels positioned anterior to native footprint and high in roof notch TT graft orientation relatively vertical to native ACL in both sagittal and coronal planes TT graft orientation vertical to native ACL in both sagittal and coronal planes AM more accurate than TT reconstructions in femoral tunnel position when comparing margin of footprint and border of articular surface of lateral femoral condyle Sagittal graft obliquity better restored with AM than TT technique compared with native, contralateral ACL TT femoral tunnels were positioned proximal to native footprint and produced more vertical grafts, whereas AM tunnels were centered and produced more horizontal grafts with better restoration of stability
been identified in TT reconstructions performed with traditional techniques.6,18,36 In this study our hypothesis that the tibial tunnel would be significantly posterior with the TT technique compared with the AM portal technique is supported by the data. In the hands of exceptionally experienced surgeons, adequate femoral tunnel positioning could be achieved but with tibial tunnels that were significantly posterior to the native tibial footprint centroid. This posterior tibial insertion ultimately resulted in a nearly 15° discrepancy in mean sagittal tunnel obliquity, with TT ACL grafts significantly vertical and AM ACL grafts restorative of native ligament sagittal obliquity. The genesis of our study was based on a concern that the tibial tunnel may be malpositioned, albeit inadvertently, to achieve satisfactory femoral tunnel position with TT ACL reconstruction. Because, with the AM technique, the femoral tunnel is drilled independent of the tibial tunnel, the latter can be optimally positioned with respect to the native tibial footprint without concern for potential compromise or technical difficulty with satisfactory preparation of the femoral tunnel. In our clinical practice, we have encountered failed ACL reconstructions in which femoral tunnel position seemed reasonable, but the tibial tunnel position used to achieve this femoral position was believed to be markedly posterior. Posterior tibial position may result in inferior control of the pivot during stability testing.7,37 We have also encountered tibial tunnels so short and oblique as to compromise fixation
1520
A. L. BOWERS ET AL.
or even the articular cartilage of the medial tibial plateau. This is consistent with work by Heming et al.5 in which they found that TT reconstructions that were centered on both femoral and tibial native insertions required tibial tunnels that were excessively short and prohibitively close to the joint line. Posterior positioning of the tibial tunnel was also historically recommended with a TT technique to minimize risk for impingement seen with vertical grafts. In our clinical experience of nearly 6 years using the AM technique, we have not experienced graft impingement to be a clinical problem, likely because of the more oblique positioning of the graft in both sagittal and coronal planes.6 This is supported by work done by Iriuchishima et al.,38 who showed no significant roof, posterior cruciate ligament, or interbundle impingement with anatomic single- or double-bundle ACL reconstruction in a cadaveric model. Examination of graft positioning and obliquity by comparison with 3D reconstructions of the native, contralateral knee shows that although both techniques capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. We found that when a tibial tunnel is drilled without the need to accommodate for subsequent femoral tunnel reaming, a more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. Although this MRI analysis of postoperative subjects does indicate improved re-creation of native anatomy by use of the AM technique compared with the TT technique, a difference in functional outcomes favoring one technique over another with respect to graft function and patient outcomes remains undetermined. We acknowledge that coronal plane obliquity was actually increased relative to the native ACL using the AM technique compared with the TT technique. It would appear that when using this single-bundle construct, a tradeoff occurs in optimizing sagittal versus coronal obliquity. Increased coronal obliquity may contribute to constraint of tibial internal rotation. In a cross-sectional study examining gait analysis after ACL reconstruction, Scanlan et al.39 found a negative correlation between the peak external knee flexion moment during walking and the coronal angle of the ACL graft; they posited that with more vertical coronal orientation, patients reduce their net quadriceps usage during walking. The clinical implications of this finding, and of the findings reported in our study, remain unknown. A weakness of this study is the inability to assess
graft position relative to individual AM and PL ACL bundles. Recent biomechanical analyses have shown that vertical “graft mismatch” connecting the AM bundle femoral insertion with the PL bundle tibial insertion has inferior kinematic behavior relative to matched reconstructions, whereas a horizontal “mismatch” connecting the PL femur with the AM tibia succeeds in limiting translation with both Lachman and pivot-shift maneuvers.20 We suspect that the configuration achieved with our AM portal ACL reconstructions mimics this PL femur–AM tibia construct, particularly given the more horizontal position achieved in the coronal plane. However, because of limitations of our MRI reconstructions, we were unable to reliably differentiate between the 2 bundles consistently throughout the length of the contralateral, native ACL. Multiple surgeons were used in this study. However, we observed that tunnel positions were consistent among surgeons using the same technique, and we believe that inclusion of multiple surgeons may serve to reduce the likelihood of bias resulting from the techniques of a single surgeon. In each case the surgeon selected the technique that he or she felt, in his or her hands, would best position the graft based on the patient’s anatomy. This should make findings more ubiquitous to the general orthopaedic community. Nonetheless, we should caution that all reconstructions evaluated were performed by high-volume, fellowship-trained sports medicine surgeons. Furthermore, AM reconstructions are performed at our institution with flexible reaming systems, and our findings may not be generalizable to AM reconstructions with standard, nonflexible reaming systems. Finally, although clear differences between techniques have been elucidated in this study by MRI, no subjective or objective clinical outcomes have been evaluated, because they are beyond the scope of this initial investigation. The true clinical relevance of these MRI measurements remains unknown and is a source of critical ongoing evaluation. CONCLUSIONS In this series AM portal femoral tunnel reaming more accurately restored native ACL anatomy than did TT reaming. Although both techniques can capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. When a tibial tunnel is drilled without the need to accommodate subse-
ACL TUNNEL POSITION AND GRAFT OBLIQUITY quent femoral tunnel reaming, more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. 16.
Acknowledgment: The authors thank the following physicians for their assistance: Jo A. Hannafin, M.D., Ph.D., Stephen J. O’Brien, M.D., M.B.A., Howard Anthony Rose, M.D., Anne Kelly, M.D., Beth Shubin Stein, M.D., Struan Coleman, M.D., Ph.D., Matthew Koff, Ph.D., and Shirley Magabo, M.D.
17.
18.
REFERENCES 1. Chhabra A, Diduch DR, Blessey PB, Miller MD. Recreating an acceptable angle of the tibial tunnel in the coronal plane in anterior cruciate ligament reconstruction using external landmarks. Arthroscopy 2004;20:328-330. 2. Howell SM, Gittins ME, Gottlieb JE, Traina SM, Zoellner TM. The relationship between the angle of the tibial tunnel in the coronal plane and loss of flexion and anterior laxity after anterior cruciate ligament reconstruction. Am J Sports Med 2001;29:567-574. 3. Dargel J, Schmidt-Wiethoff R, Fischer S, Mader K, Koebke J, Schneider T. Femoral bone tunnel placement using the transtibial tunnel or the anteromedial portal in ACL reconstruction: A radiographic evaluation. Knee Surg Sports Traumatol Arthrosc 2009;17:220-227. 4. Giron F, Cuomo P, Edwards A, Bull AM, Amis AA, Aglietti P. Double-bundle “anatomic” anterior cruciate ligament reconstruction: A cadaveric study of tunnel positioning with a transtibial technique. Arthroscopy 2007;23:7-13. 5. Heming JF, Rand J, Steiner ME. Anatomical limitations of transtibial drilling in anterior cruciate ligament reconstruction. Am J Sports Med 2007;35:1708-1715. 6. Steiner ME, Battaglia TC, Heming JF, Rand JD, Festa A, Baria M. Independent drilling outperforms conventional transtibial drilling in anterior cruciate ligament reconstruction. Am J Sports Med 2009;37:1912-1919. 7. Bedi A, Musahl V, Steuber V, et al. Transtibial versus anteromedial portal reaming in anterior cruciate ligament reconstruction: An anatomic and biomechanical evaluation of surgical technique. Arthroscopy 2011;27:380-390. 8. Bottoni CR. Anterior cruciate ligament femoral tunnel creation by use of anteromedial portal. Arthroscopy 2008;24:1319. 9. Cain EL Jr, Clancy WG Jr. Anatomic endoscopic anterior cruciate ligament reconstruction with patella tendon autograft. Orthop Clin North Am 2002;33:717-725. 10. Harner CD, Honkamp NJ, Ranawat AS. Anteromedial portal technique for creating the anterior cruciate ligament femoral tunnel. Arthroscopy 2008;24:113-115. 11. Lubowitz JH. Anteromedial portal technique for the anterior cruciate ligament femoral socket: Pitfalls and solutions. Arthroscopy 2009;25:95-101. 12. Shelbourne KD, Patel DV. ACL reconstruction using the autogenous bone-patellar tendon-bone graft: Open two-incision technique. Instr Course Lect 1996;45:245-252. 13. Steiner ME, Murray MM, Rodeo SA. Strategies to improve anterior cruciate ligament healing and graft placement. Am J Sports Med 2008;36:176-189. 14. Abebe ES, Moorman CT III, Dziedzic TS, et al. Femoral tunnel placement during anterior cruciate ligament reconstruction: An in vivo imaging analysis comparing transtibial and 2-incision tibial tunnel-independent techniques. Am J Sports Med 2009; 37:1904-1911. 15. Brophy RH, Voos JE, Shannon FJ, et al. Changes in the length
19.
20.
21.
22.
23. 24. 25.
26.
27.
28. 29. 30.
31.
1521
of virtual anterior cruciate ligament fibers during stability testing: A comparison of conventional single-bundle reconstruction and native anterior cruciate ligament. Am J Sports Med 2008;36:2196-2203. Forsythe B, Kopf S, Wong AK, et al. The location of femoral and tibial tunnels in anatomic double-bundle anterior cruciate ligament reconstruction analyzed by three-dimensional computed tomography models. J Bone Joint Surg Am 2010;92: 1418-1426. Gavriilidis I, Motsis EK, Pakos EE, Georgoulis AD, Mitsionis G, Xenakis TA. Transtibial versus anteromedial portal of the femoral tunnel in ACL reconstruction: A cadaveric study. Knee 2008;15:364-367. Kopf S, Forsythe B, Wong AK, et al. Nonanatomic tunnel position in traditional transtibial single-bundle anterior cruciate ligament reconstruction evaluated by three-dimensional computed tomography. J Bone Joint Surg Am 2010;92:14271431. Miller MD, Gerdeman AC, Miller CD, et al. The effects of extra-articular starting point and transtibial femoral drilling on the intra-articular aperture of the tibial tunnel in ACL reconstruction. Am J Sports Med 2010;38:707-712. Pearle AD, Shannon FJ, Granchi C, Wickiewicz TL, Warren RF. Comparison of 3-dimensional obliquity and anisometric characteristics of anterior cruciate ligament graft positions using surgical navigation. Am J Sports Med 2008;36:15341541. Alentorn-Geli E, Lajara F, Samitier G, Cugat R. The transtibial versus the anteromedial portal technique in the arthroscopic bone-patellar tendon-bone anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc 2010;18:1013-1037. Hantes ME, Zachos VC, Liantsis A, Venouziou A, Karantanas AH, Malizos KN. Differences in graft orientation using the transtibial and anteromedial portal technique in anterior cruciate ligament reconstruction: A magnetic resonance imaging study. Knee Surg Sports Traumatol Arthrosc 2009;17:880-886. Bedi A, Altchek DW. The “footprint” anterior cruciate ligament technique: An anatomic approach to anterior cruciate ligament reconstruction. Arthroscopy 2009;25:1128-1138. Asano T, Akagi M, Tanaka K, Tamura J, Nakamura T. In vivo three-dimensional knee kinematics using a biplanar imagematching technique. Clin Orthop Relat Res 2001:157-166. Loh JC, Fukuda Y, Tsuda E, Steadman RJ, Fu FH, Woo SL. Knee stability and graft function following anterior cruciate ligament reconstruction: Comparison between 11 o’clock and 10 o’clock femoral tunnel placement. 2002 Richard O’Connor Award paper. Arthroscopy 2003;19:297-304. Musahl V, Plakseychuk A, VanScyoc A, et al. Varying femoral tunnels between the anatomical footprint and isometric positions: Effect on kinematics of the anterior cruciate ligament-reconstructed knee. Am J Sports Med 2005;33:712-718. Yamamoto Y, Hsu WH, Woo SL, Van Scyoc AH, Takakura Y, Debski RE. Knee stability and graft function after anterior cruciate ligament reconstruction: A comparison of a lateral and an anatomical femoral tunnel placement. Am J Sports Med 2004;32:1825-1832. Lee MC, Seong SC, Lee S, et al. Vertical femoral tunnel placement results in rotational knee laxity after anterior cruciate ligament reconstruction. Arthroscopy 2007;23:771-778. Morgan CD, Kalman KR, Grawl DM. Definitive landmarks for reproducible tibial tunnel placement in anterior cruciate ligament reconstruction. Arthroscopy 1995;11:275-288. Piasecki DP, Bach BR Jr, Espinoza Orias AA, Verma NN. Anterior cruciate ligament reconstruction: Can anatomic femoral placement be achieved with a transtibial technique? Am J Sports Med 2011;39:1306-1315. Arnold MP, Kooloos J, van Kampen A. Single-incision tech-
1522
32.
33.
34. 35.
A. L. BOWERS ET AL.
nique misses the anatomical femoral anterior cruciate ligament insertion: A cadaver study. Knee Surg Sports Traumatol Arthrosc 2001;9:194-199. Ahn JH, Lee SH, Yoo JC, Ha HC. Measurement of the graft angles for the anterior cruciate ligament reconstruction with transtibial technique using postoperative magnetic resonance imaging in comparative study. Knee Surg Sports Traumatol Arthrosc 2007;15:1293-1300. Stanford FC, Kendoff D, Warren RF, Pearle AD. Native anterior cruciate ligament obliquity versus anterior cruciate ligament graft obliquity: An observational study using navigated measurements. Am J Sports Med 2009;37:114-119. Giron F, Buzzi R, Aglietti P. Femoral tunnel position in anterior cruciate ligament reconstruction using three techniques. A cadaver study. Arthroscopy 1999;15:750-756. Albuquerque RF, Amatuzzi MM, Pacheco AP, Angelini FJ, Campos O Jr. Positioning of the femoral tunnel for arthroscopic reconstruction of the anterior cruciate ligament: Comparative study of 2 techniques. Clinics (Sao Paulo) 2007;62:613-618.
36. Purnell ML, Larson AI, Clancy W. Anterior cruciate ligament insertions on the tibia and femur and their relationships to critical bony landmarks using high-resolution volume-rendering computed tomography. Am J Sports Med 2008;36:20832090. 37. Voos JE, Musahl V, Maak TG, Wickiewicz TL, Pearle AD. Comparison of tunnel positions in single-bundle anterior cruciate ligament reconstructions using computer navigation. Knee Surg Sports Traumatol Arthrosc 2010;18:12821289. 38. Iriuchishima T, Tajima G, Ingham SJ, Shen W, Smolinski P, Fu FH. Impingement pressure in the anatomical and nonanatomical anterior cruciate ligament reconstruction: A cadaver study. Am J Sports Med 2010;38:1611-1617. 39. Scanlan SF, Blazek K, Chaudhari AM, Safran MR, Andriacchi TP. Graft orientation influences the knee flexion moment during walking in patients with anterior cruciate ligament reconstruction. Am J Sports Med 2009;37:2173-2178.
Purpose: Using 3-dimensional high-resolution magnetic resonance imaging (MRI), we sought to compare femoral and tibial tunnel position and resultant graft obliquity with single-bundle anterior cruciate ligament (ACL) reconstruction using transtibial (TT) or anteromedial (AM) portal femoral tunnel reaming techniques. Methods: Thirty patients were prospectively enrolled after primary, autogenous bone–patellar tendon– bone ACL reconstruction by 2 groups of high-volume, fellowship-trained sports medicine surgeons. With the TT technique, an external starting point was used to maximize graft obliquity and femoral footprint capture. By use of high-resolution MRI and imaging analysis software, bilateral 3-dimensional knee models were created, mirrored, and superimposed. Differences between centroids for each femoral and tibial insertion, as well as corresponding ACL/graft obliquity, were evaluated with paired t tests and 2-sided Mann-Whitney nonparametric tests, with P ⬍ .05 defined as significant. Results: No significant differences were observed between groups in position of reconstructed femoral footprints. However, on the tibial side, AM centroids averaged 0.8 ⫾ 1.9 mm anterior to native ACL centroids, whereas the TT group centered 5.23 ⫾ 2.4 mm posterior to native ACL centroids (P ⬍ .001). Sagittal obliquity was closely restored with the AM technique (mean, 52.2° v 53.5° for native ACL) but was significantly more vertical (mean, 66.9°) (P ⫽ .0001) for the TT group. Conclusions: In this clinical series, AM portal femoral tunnel reaming more accurately restored native ACL anatomy than the TT technique. Although both techniques can capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. When a tibial tunnel is drilled without the need to accommodate subsequent femoral tunnel reaming, more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. Level of Evidence: Level III, retrospective comparative study.
E
ndoscopic anterior cruciate ligament (ACL) reconstruction with transtibial (TT) preparation of a femoral socket has been performed for nearly 3 decades with good clinical success. With improved un-
derstanding of the complex anatomy and function of the native ligament, however, modifications to the TT technique have been suggested to improve femoral tunnel position and restore the native ligament foot-
From the Sports Medicine and Shoulder Service, Hospital for Special Surgery, New York, New York; and MedSport, Department of Orthopaedic Surgery, University of Michigan (A.B.), Ann Arbor, Michigan, U.S.A. Supported by the Institute for Sports Medicine Research. The authors report no conflict of interest. Received March 21, 2011; accepted July 6, 2011. Address correspondence to David W. Altchek, M.D., 535 E 70th St, New York, NY 10021, U.S.A. E-mail: [email protected] © 2011 by the Arthroscopy Association of North America 0749-8063/11180/$36.00 doi:10.1016/j.arthro.2011.07.007
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 27, No 11 (November), 2011: pp 1511-1522
1511
1512
A. L. BOWERS ET AL.
prints and graft obliquity.1,2 Nonetheless, questions have been raised regarding the ability of the TT technique to accurately and safely re-create the correct femoral and tibial insertion sites.3-6 In the past decade, many investigators have advocated preparation of the femoral tunnel independently through an anteromedial (AM) portal or 2-incision technique, thereby avoiding the constraints of tibial tunnel drilling.7-13 Preliminary navigation studies in cadavers have shown improved tunnel position and graft kinematics with techniques that use independent femoral tunnel drilling with ACL reconstruction.6,7,14-20 Clinical comparisons of the 2 techniques, however, are limited.21 A previous magnetic resonance imaging (MRI) study of graft obliquity by Hantes et al.22 has been limited to use of isolated 2-dimensional slices of tunnels in an attempt to characterize a 3-dimensional (3D) structure. Recently established was a new, validated method for tunnel evaluation using 3D MRI models that compare a mirror-image overlay of the center of the femoral tunnel aperture with the center of the native ACL attachment on the contralateral, native knee.14 Such models were generated by Abebe et al.14 to show that independent drilling with an outside-in guide pin and use of a retrograde drill (Arthrex, Naples, FL) resulted in a femoral tunnel positioned significantly closer to the center of the native ACL whereas the TT technique resulted in femoral tunnels often centered outside of the native footprint. We favor the use of this 3D technique because it allows for direct comparison with the patient’s own native ACL anatomy and determination of the exact fiber insertion rather than simply the tunnel position. This technique minimizes errors due to signal averaging and through-plane resolution when evaluating ACL anatomy. To our knowledge, however, there are no clinical reports in the published literature using 3D MRI to directly compare TT and AM portal drilling techniques. The specific aim was to compare femoral and tibial tunnel position and resultant graft obliquity achieved with single-bundle ACL reconstruction using a TT or AM portal reaming technique. We evaluated 3 main hypotheses: 1. Femoral tunnel position would be significantly improved and closer to the native femoral ACL attachment with an AM portal compared with a TT reaming technique. 2. Tibial tunnel position would be significantly posterior relative to the native tibial ACL footprint with the TT technique compared with the AM portal technique.
3. The resultant ACL graft would be more vertical in both the sagittal and coronal planes with the TT technique compared with the AM portal technique. METHODS Patient Selection and Study Design This study was approved by our Institutional Review Board (Hospital for Special Surgery, No. 29116). Inclusion criteria comprised patients aged 16 to 40 years who underwent elective, primary ACL reconstruction with autograft bone–patellar tendon– bone. Exclusion criteria were previous surgery on either the ipsilateral or contralateral knee, multiligamentous reconstruction, concomitant bony procedures that might distort the bony anatomy (osteotomy, meniscal transplant, osteoarticular allograft or autograft), patients unwilling or unable to undergo MRI, or patients with contraindications to MRI (e.g., pacemaker or certain implants). Subjects were identified prospectively, and data were gathered in a cross-sectional manner during the postoperative period. Fifteen subjects were enrolled for each technique (TT and AM). Operative notes for each were collected and scrutinized to (1) confirm the method used for femoral tunnel preparation (TT or AM) and (2) confirm that no concomitant ligamentous or bony procedure had been performed. A diagram of enrollment and protocol is depicted in Fig 1. ACL reconstructions were performed by 1 of 8 highvolume, fellowship-trained sports medicine surgeons who attempted to optimize femoral tunnel position and
FIGURE 1. Subject selection criteria, enrollment, and protocol. (B-PT-B, bone–patellar tendon– bone).
ACL TUNNEL POSITION AND GRAFT OBLIQUITY re-create native graft obliquity. TT reconstructions were performed with standard rigid reamers for each tunnel. The tibial tunnel was positioned within a portion of the tibial footprint, centered between the tibial eminences, while protecting the anterior horn of the lateral meniscus. Coronal angulation of the guide and resultant external tibial positioning were ultimately selected to allow anticipated closest access to the anatomic center of the femoral footprint through appropriate rotation of an endofemoral guide. Notchplasty was performed as needed to improve visualization, and while viewing with a 30° arthroscope, the surgeon passed a rigid guide pin and reamer through the center of the native femoral ACL footprint for preparation of the femoral tunnel. Our AM technique has been described previously.23 In brief, AM reconstructions were performed with positioning of a tibial guide and resultant tunnel within the native ACL footprint centered both anteroposteriorly and mediolaterally, without reference to surrounding anatomic landmarks. Femoral tunnel preparation was performed through an AM portal positioned immediately above the medial meniscus while allowing direct access to the lateral notch. A gentle notchplasty was performed to improve visualization. While viewing with a 70° arthroscope, the surgeon marked the center of the anatomic footprint with an awl, and with the knee held in approximately 115° of flexion, a flexible guidewire was passed from this point in a superolateral direction to maximize tunnel length. A flexible reaming system (Smith & Nephew, Andover, MA) was then used for tunnel preparation. Three-Dimensional MRI Reconstructions Magnetic resonance images were obtained between 5 and 25 weeks postoperatively. All scans were performed on a clinical 3-T scanner (GE Healthcare, Waukesha, WI) by use of an 8-channel phased-array transmit/receive knee coil. A magnetic resonance image of the contralateral knee was also acquired to serve as a control data set, and there was no history of ACL injury on the unaffected side. Each knee was placed in standardized slight flexion and external rotation during scanning. Morphologic changes of the joint were assessed from 2-dimensional fast-spin echo images acquired along 3 anatomic planes (echo time, 25 to 30 milliseconds; repetition time, 4,000 to 6,000 milliseconds; echo train length, 8 to 16; bandwidth, ⫾62.5 kHz over entire frequency range; acquisition matrix, 512 ⫻ 416 to 512 ⫻ 481; number of excitations, 1 to 2; field of view, 15 to 16 cm; slice thickness, 3.5 mm with no gap).
1513
In addition, a 3D fast-spin echo technique was used to acquire a data set with near-isotropic voxels (echo time, 36 milliseconds; repetition time, 2,500 milliseconds; echo train length, 64; bandwidth, ⫾41.67 kHz over entire frequency range; acquisition matrix, 256 ⫻ 256; number of excitations, 0.5; field of view, 18 cm; slice thickness, 0.6 mm; scan time, 7 minutes). Tissue contrast was provided to ensure differential contrast between ACL graft, bone, fluid, and cartilage. With the operator blinded to the technique used, the magnetic resonance images were segmented and 3D models were then created by use of Mimics 13.1 software (Materialise, Leuven, Belgium). The 3D models were exported to Geomagic Studio 11 (Geomagic, Research Triangle Park, NC). A mirror image of the left side was created. Multiple, corresponding anatomic points between the 2 femurs were registered, and the mirrored left side was rotated and translated for a best-fit overlay onto the right femur. The rotational and translational transformation was then applied for the left-side ACL or graft femoral footprint. The same technique of registration, overlay, and footprint transformation was then applied to the tibial side. Femoral and tibial sides were registered individually to avoid any error that could be generated if there were any differences between sides in terms of sagittal attitude (slightly flexed posture on one side v full extension on the other side at the time the MRI sequence was acquired). Determination of Centroids and Graft Obliquity The centroids of the femoral and tibial insertions of both the ACL and graft were then determined. To describe the positions of the reconstructed centroids relative to native centroids, and to compare between subjects, tunnel positioning was assessed relative to reproducible anatomic landmarks on both sides of the joint. On the femoral side, best-fit spheres were identified for the contour of each of the medial and lateral distal femoral condyles (Fig 2), and the centers of these spheres were used to determine the geometric center axis, as described and validated previously by Asano et al.24 This provided a medial-lateral reference axis. The anterior-posterior axis was calculated perpendicular to the medial-lateral axis. The proximaldistal axis was determined by calculating 2 centroids of cross sections of the periosteal surface of the femur 30 mm apart up the femoral shaft. This established a coordinate system by which the distance between the centroids of the reconstructed graft and native ACL could be described in both the anterior-posterior and
1514
A. L. BOWERS ET AL.
FIGURE 2. Best-fit spheres are identified for the contour of each of the medial and lateral distal femoral condyles. The medial-lateral reference axis was determined by the geographic centers of the spheres, as described and validated previously by Asano et al.24 The anterior-posterior axis was calculated perpendicular to the medial-lateral axis.
proximal-distal directions, measured in millimeters (Fig 3). On the tibial side, the 3D model was rotated so that the articular surface could be viewed en face. The medialand lateral-most points were registered. Next, a best-fit plane was created to match the surface of the plateau medially and laterally. The anterior-posterior axis was determined perpendicular to the medial-lateral axis in this plane. The distance of the tibial centroid of the reconstructed graft could then be described in both anterior-posterior and medial-lateral directions relative to the native tibial footprint, again measured in millimeters (Fig 3).
FIGURE 3. Sample reconstructions for the femur (left) and tibia (right). Distances were measured (in millimeters) between the native and reconstructed centroids in the anterior-posterior and proximal-distal directions for the femur and in the medial-lateral and anterior-posterior directions for the tibia. The native footprints are depicted in green and the reconstructed footprints in red.
Next, the obliquities of the native ACL and reconstructed grafts were measured in the sagittal and coronal planes by use of Pro/Engineer Wildfire 4.0 (Parametric Technology, Needham, MA). These were determined by the angle subtended by a line extending from the centers of the femoral and tibial footprints and the best-fit plane of the tibial surface (Fig 4). Footprint Overlap On the tibial side, the extent of overlap of the surface area for the graft insertion and the native footprint was calculated. By use of ImageJ 1.40g
ACL TUNNEL POSITION AND GRAFT OBLIQUITY
1515
FIGURE 4. Sample measurements for coronal (left) and sagittal (right) graft obliquities. The native footprint and ACL obliquity are depicted in green and the reconstructed graft in red.
(National Institutes of Health, Bethesda, MD), the percent overlap between the native and reconstructed sides was then calculated to determine what portion of the graft insertion fell within the native footprint. Statistical Analysis A sample size of 15 was determined for each group (AM and TT) to achieve 80% power to detect a difference of 3 mm from the center of the native ACL between the 2 approaches, with a conservatively estimated SD of ⫾3 mm and a significance level of .050, as well as adjustment for a 2-sided Mann-Whitney nonparametric test. Graft placement relative to the native ACL was then compared between techniques by use of paired t tests. Native ACL obliquity in the sagittal and coronal planes between the TT and AM groups was compared by use of unpaired t tests to confirm that there were no differences between controls for each group. A graft obliquity difference of 10° from the native ACL in the sagittal plane was considered clinically significant. Notably, 4 scans were segmented and processed in duplicate to verify accuracy; each yielded measurements reproducible within 0.8 mm and 1.0°. Because of the time-intensive nature of the segmentation and processing (approximately 8 hours per subject), the remaining scans were not analyzed in duplicate.
RESULTS We contacted 35 consecutive subjects meeting the inclusion/exclusion criteria for participation; 5 (2 AM and 3 TT) declined participation, citing the time required to sit through bilateral knee MRI. Thus 30 subjects were consecutively enrolled, 15 in the AM group and 15 in the TT group. There were 16 male and 14 female patients. Of the reconstructed knees, 19 were right sided and 11 left sided. The mean age at reconstruction was 24 years (range, 16 to 38 years). The mean length of time from reconstruction to postoperative MRI was 12 weeks (range, 5 to 25 weeks). Data for each subject are depicted in Table 1. Centroid Position On the femoral side, centroids from the AM portal group averaged 1.97 mm anterior (range, 1.60 mm posterior to 4.52 mm anterior), 1.91 mm distal (range, 2.08 mm proximal to 4.20 mm distal), and 2.89 mm (range, 0.50 to 5.90 mm) in greatest distance (hypotenuse) from the matched-pair, native ACL femoral centroids. Centroids prepared with the TT technique averaged 3.02 mm anterior (range, 2.60 mm posterior to 9.91 mm anterior), 2.46 mm distal (range, 2.77 mm proximal to 6.53 mm distal), and 2.15 mm (range,
1516
A. L. BOWERS ET AL. TABLE 1.
Data Collected for 30 Study Participants Difference in Graft Obliquity (°) Relative to Paired, Native ACL
Footprint Centroid Position Femoral
Subject No. Anteromedial group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 TT group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Tibial Maximum Distance (mm)
Direction Relative to Paired, Native ACL
Sagittal
Coronal
Tibial Footprint: Graft Overlap With Native Footprint (%)
Anterodistal Anterodistal Anterodistal Posterodistal Posterodistal Posteroproximal Posteroproximal Posteroproximal Anterodistal Anterodistal Anterodistal Posterodistal Anteroproximal Anterodistal Anterodistal
5.05 4.22 0.76 2.59 2.08 3.39 6.87 3.34 2.42 1.79 2.20 5.08 0.42 1.20 2.78
Anteromedial Anteromedial Posteromedial Anteromedial Posteromedial Posteromedial Anteromedial Anteromedial Anterolateral Posteromedial Posteromedial Posteromedial Anteromedial Anteromedial Anteromedial
⫺0.2 ⫺0.1 ⫺5 7.8 4.7 ⫺5.9 7.7 2.1 0.1 0.5 ⫺2.7 ⫺0.1 ⫺0.5 0.9 ⫺2.3
13.3 20 ⫺0.7 15.7 12.9 10.4 11.9 3 ⫺0.5 6.6 11 14.2 0.3 2.7 12
72.1 56.3 65.3 83.6 63.7 67.5 48.3 42.4 79.5 52.9 55.6 63.4 77.3 89.8 76.6
Anterodistal Posteroproximal Anterodistal Anteroproximal Anteroproximal Anteroproximal Anterodistal Anterodistal Posterodistal Anterodistal Anterodistal Anterodistal Anterodistal Anterodistal Anterodistal
5.67 6.27 11.00 9.68 2.61 4.29 5.34 2.04 7.09 7.56 5.21 3.06 5.88 6.70 4.47
Posteromedial Posteromedial Posterolateral Posteromedial Posteromedial Posterolateral Posteromedial Anteromedial Posteromedial Anteromedial Posteromedial Posteromedial Posteromedial Posteromedial Posterolateral
⫺11.4 ⫺11.2 ⫺27.5 ⫺17.3 ⫺15.8 ⫺13.1 ⫺12.9 ⫺3 ⫺8.1 8.9 ⫺25.2 ⫺3.4 ⫺16.7 ⫺23.1 ⫺7.5
13.7 8.8 ⫺0.7 13 3.7 ⫺4.1 6.7 ⫺2.3 7.5 13.4 6.3 4.7 ⫺10.8 2.9 ⫺2.4
29.4 36 5.6 6 44.4 57.1 55.6 74.5 39.1 29.1 42.2 43.1 39.2 19 48.9
Maximum Distance (mm)
Direction Relative to Paired, Native ACL
0.50 5.90 5.62 3.69 2.23 1.70 1.06 2.08 5.00 1.14 3.71 2.52 1.48 1.63 5.11 1.65 1.46 10.72 2.19 4.56 2.60 3.88 3.37 5.18 0.83 11.19 4.23 4.08 3.52 2.80
0.83 to 11.19 mm) in greatest distance from the paired, native ACL femoral centroids. There were no statistically significant differences between the 2 groups with respect to distance from the native centroid in the anterior-posterior (P ⫽ .206) or proximaldistal (P ⫽ .344) direction or for the hypotenuse (P ⫽ .176) (Table 2). Femoral data points are depicted in Fig 5. On the tibial side, centroids from the AM portal group averaged 1.31 mm anterior (range, 1.52 mm posterior to 6.09 mm anterior), 2.41 mm medial (range, 5.01 mm medial to 2.07 mm lateral), and 2.83
mm (range, 0.42 to 6.87 mm) in greatest distance (hypotenuse) from the paired, native ACL tibial centroids. TT centroids averaged 5.23 mm posterior (range, 11.00 mm posterior to 6.54 mm anterior), 1.91 mm medial (range, 5.45 mm medial to 1.27 mm lateral), and 5.75 mm (range, 2.04 to 11.00 mm) in greatest distance from paired, native ACL tibial centroids. These findings were significantly different for the anterior-posterior plane (P ⬍ .001) and greatest distance (P ⫽ .001) but not for the medial-lateral plane (P ⫽ .344) (Table 2). Tibial data points are depicted in Fig 6.
ACL TUNNEL POSITION AND GRAFT OBLIQUITY
1517
TABLE 2. Comparison of Tunnel Positions Between AM Portal and TT Femoral Tunnel Reaming Techniques Insertion Direction Femur Anterior/posterior Proximal/distal Hypotenuse Tibia Anterior/posterior Medial/lateral Hypotenuse
Technique
Mean Distance (SD) (mm)
AM TT AM TT AM TT
1.97 (1.52) 3.05 (2.87) 1.91 (1.35) 2.46 (1.73) 2.89 (1.80) 4.15 (3.02)
.206
AM TT AM TT AM TT
1.31 (1.54) 5.23 (2.40) 2.41 (1.40) 1.91 (1.44) 2.83 (1.84) 5.75 (2.36)
⬍.001
P Value
.344 .176
.344 .001
Graft Obliquity In the sagittal plane, native ACL obliquity averaged 53.5°, and no significant difference was observed between the native ACL obliquity in the TT group (52.6°) and AM group (54.4°) (P ⫽ .430). Mean graft obliquity in the AM portal group was 52.2° (range, 44.3° to 60.3°) and was not significantly different from that of the paired, native ACL (P ⫽ .653). TT-reconstructed graft obliquity, however, averaged 66.9° (range, 45.5° to 78.4°) and was
FIGURE 6. Tibial position of reconstructed footprint centroids relative to native footprint centroids. A representative composite of native tibial footprints is marked in green, with its centroid marked in gray. Centroids of grafts reconstructed with the TT technique are marked in red, and those with the AM portal technique are marked in blue. The TT technique resulted in a position significantly posterior to centroids for both native ACLs and the AM portal technique.
significantly less oblique than that of the native ACL (P ⫽ .0001) (Table 3). In the coronal plane, native ACL obliquity averaged 76.1°, and no difference was observed between native ACL obliquity in the TT group (77.1°) and AM group (75.0°) (P ⫽ .326). AM-reconstructed graft coronal obliquity averaged 66.1° (range, 58.2° to 77.5°) and was significantly more oblique than that of the paired, native ACL (P ⬍ .0001). TT-reconstructed graft obliquity averaged 73.0° (range, 64.6° to 79.1°) and
TABLE 3. Comparison of Graft Obliquities Between AM Portal and TT Femoral Tunnel Reaming Techniques Obliquity Sagittal Native ACL AM TT
FIGURE 5. Femoral position of reconstructed footprint centroids relative to native footprint centroids. A representative composite of native femoral footprints is marked in green, with its centroid marked in gray. Centroids of grafts reconstructed with the TT technique are marked in red, and those with the AM portal technique are marked in blue. No significant differences were seen among centroid positions between these techniques.
Coronal Native ACL AM TT
Technique
Mean Angle (SD) (°)
AM TT Reconstructed Native Reconstructed Native
52.6 (4.33) 54.4 (7.15) 52.2 (5.07) 52.6 (4.33) 66.9 (9.11) 54.4 (7.15)
.430
AM TT Reconstructed Native Reconstructed Native
75.0 (5.67) 77.1 (5.62) 66.1 (4.69) 75.0 (5.67) 73.0 (4.50) 77.1 (5.62)
.326
P Value
.653 .0001
⬍.0001 .045
1518
A. L. BOWERS ET AL.
FIGURE 7. Maximum, mean, and minimum tibial footprint fill for AM portal and TT techniques. Native footprints are depicted in green, and reconstructed footprints are shown in red. Tibial graft insertion prepared with the AM technique had a mean of 66.3% overlap with the native ACL tibial footprint. For the TT technique, the overlap was 38.0% (P ⬍ .0001).
also was somewhat greater than that of paired, native ACLs (P ⫽ .045) (Table 3). Tibial Footprint Fill With respect to footprint fill, tibial graft insertion prepared with the AM technique had a 66.3% overlap with the native ACL. The portion that did not overlap fell predominantly medial to the native insertion. For the TT technique, the overlap was 38.0% (P ⬍ .0001); the remaining 62% fell predominantly posterior and slightly medial compared with the native footprint (Fig 7). DISCUSSION Independent drilling of the femoral tunnel through an AM portal compared with TT reaming has been theorized to better restore native anatomy in ACL reconstruction. We hypothesized that reconstructions performed with AM portal reaming would yield more accurate femoral position, tibial position, and graft obliquity. In this cross-sectional MRI study, we found that both techniques performed equally well with respect to femoral tunnel position. However, the AM portal technique resulted in more anatomic tibial tunnel position and sagittal graft obliquity. True “anatomic” ACL reconstruction remains elusive because of the complex native anatomy with differing fiber lengths. Nonetheless, ACL reconstruction with attachment points that more closely matched native ligament attachment location has been repeatedly shown to better reproduce knee kinematics in both cadaveric and in vivo models6,13,15,25-27 and is theorized to improve clinical outcomes. The surgical technique by which to prepare tibial and femoral sockets while respecting the native ACL anatomy, however, is controversial. Whereas some authors have recommended preparation of the femoral tunnel using a TT technique, others have advocated for indepen-
dent femoral tunnel drilling through an AM arthroscopic portal. Endoscopic reconstruction techniques have evolved in an effort to increase graft obliquity, because the relatively vertical graft prepared with traditional TT techniques succeeded in limiting anterior translation but not necessarily the rotatory instability seen in ACL insufficiency.25,28 Traditional TT techniques as described by Morgan et al.29 suggested use of internal landmarks to select a relatively posterior position on the tibia, resulting in femoral tunnel position in the proximal portion of the femoral footprint. Increasing the coronal obliquity2 or using external landmarks to select a more medial starting point1 in preparation of the tibial tunnel has been suggested to improve subsequent positioning of the femoral tunnel closer to the native anatomic ACL insertion on the femur. In navigated cadaveric specimens, femoral tunnels near the center of the femoral insertion have been shown to be achievable through a TT technique; however, these required meticulous tunnel placement “at the limits of practical,” aperture beveling, and the acceptance of some graft-tunnel mismatch for autograft bone– patellar tendon– bone grafts.30 Despite these modifications, studies have repeatedly shown that tunnel positioning remains compromised with TT ACL reconstruction techniques.3,4,31-33 Brophy et al.15 and Pearle et al.20 reported that the traditional, arthroscopic TT technique predisposed to a “mismatch” graft position from the posterolateral (PL) tibial footprint to the AM femoral footprint. Heming et al.5 found that a TT technique could produce tunnels centered in both the tibial and femoral footprints only if a starting point prohibitively close to the joint line with a correspondingly short tibial tunnel is used. Meanwhile, direct comparisons of femoral tunnel position and graft obliquity after TT or AM portal ACL reconstruction are growing in the literature. In both cadaveric6,17 and in vivo3,22 studies, femoral tun-
ACL TUNNEL POSITION AND GRAFT OBLIQUITY TABLE 4.
1519
Relevant Historical Analyses of TT Reconstructions Compared With Native ACL and AM Reconstructions
Study
Analysis
Arnold et al.31
Cadaveric arthroscopy and dissection
Giron et al.4
Cadaveric dissection and plain radiographs
Dargel et al.3
Clinical reconstruction with plain radiographs
Ahn et al.32
Clinical reconstruction with 2-dimensional MRI
Stanford et al.33 Gavriilidis et al.17
Computer navigation of cadavers v clinical reconstructions Cadaveric arthroscopy and dissection
Hantes et al.22
Clinical reconstruction with 2-dimensional MRI
Steiner et al.6
Cadaveric navigation and kinematics
nel position and graft obliquity have been shown to be more closely replicated with AM compared with TT techniques. Relevant historical studies are summarized in brief in Table 4. In our series TT reconstructions were performed by high-volume (50 to ⬎150 ACL reconstructions per year), fellowship-trained sports medicine surgeons. The goal was anatomic restoration of the native ACL footprint and ligament obliquity in all cases. We found that the femoral tunnel achieved with a TT technique was similar in position to that created with AM techniques. Our results were similar to those seen in a select few publications.34,35 Giron et al.34 reported that both TT and AM techniques were able to re-create a reference hole made just deep to the native ACL AM bundle insertion in cadaveric knees. Albuquerque et al.35 compared 10 cadaveric reconstructions performed through a TT technique with 10 performed using AM drilling. In 3 outcomes measured (posterior wall thickness, tunnel position at notch, and tunnel inclination in relation to femoral axis), no differences were observed between the 2 techniques. Contrary to our first hypothesis, it appears that improved femoral tunnel position is possible with a TT technique. However, in an attempt to produce an anatomic reconstruction, the tibial tunnel position is also of great importance, because we recognize that the positions of both tunnels ultimately determine graft obliquity with a TT technique. Posterior placement of the tibial tunnel relative to the native tibial footprint has
Pertinent Findings Using a correctly placed tibial tunnel, an anatomic attachment on the femur could not be achieved through a TT technique Impossible to restore anatomic femoral ACL origin with TT drilling technique TT drilling resulted in femoral tunnels positioned anterior to native footprint and high in roof notch TT graft orientation relatively vertical to native ACL in both sagittal and coronal planes TT graft orientation vertical to native ACL in both sagittal and coronal planes AM more accurate than TT reconstructions in femoral tunnel position when comparing margin of footprint and border of articular surface of lateral femoral condyle Sagittal graft obliquity better restored with AM than TT technique compared with native, contralateral ACL TT femoral tunnels were positioned proximal to native footprint and produced more vertical grafts, whereas AM tunnels were centered and produced more horizontal grafts with better restoration of stability
been identified in TT reconstructions performed with traditional techniques.6,18,36 In this study our hypothesis that the tibial tunnel would be significantly posterior with the TT technique compared with the AM portal technique is supported by the data. In the hands of exceptionally experienced surgeons, adequate femoral tunnel positioning could be achieved but with tibial tunnels that were significantly posterior to the native tibial footprint centroid. This posterior tibial insertion ultimately resulted in a nearly 15° discrepancy in mean sagittal tunnel obliquity, with TT ACL grafts significantly vertical and AM ACL grafts restorative of native ligament sagittal obliquity. The genesis of our study was based on a concern that the tibial tunnel may be malpositioned, albeit inadvertently, to achieve satisfactory femoral tunnel position with TT ACL reconstruction. Because, with the AM technique, the femoral tunnel is drilled independent of the tibial tunnel, the latter can be optimally positioned with respect to the native tibial footprint without concern for potential compromise or technical difficulty with satisfactory preparation of the femoral tunnel. In our clinical practice, we have encountered failed ACL reconstructions in which femoral tunnel position seemed reasonable, but the tibial tunnel position used to achieve this femoral position was believed to be markedly posterior. Posterior tibial position may result in inferior control of the pivot during stability testing.7,37 We have also encountered tibial tunnels so short and oblique as to compromise fixation
1520
A. L. BOWERS ET AL.
or even the articular cartilage of the medial tibial plateau. This is consistent with work by Heming et al.5 in which they found that TT reconstructions that were centered on both femoral and tibial native insertions required tibial tunnels that were excessively short and prohibitively close to the joint line. Posterior positioning of the tibial tunnel was also historically recommended with a TT technique to minimize risk for impingement seen with vertical grafts. In our clinical experience of nearly 6 years using the AM technique, we have not experienced graft impingement to be a clinical problem, likely because of the more oblique positioning of the graft in both sagittal and coronal planes.6 This is supported by work done by Iriuchishima et al.,38 who showed no significant roof, posterior cruciate ligament, or interbundle impingement with anatomic single- or double-bundle ACL reconstruction in a cadaveric model. Examination of graft positioning and obliquity by comparison with 3D reconstructions of the native, contralateral knee shows that although both techniques capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. We found that when a tibial tunnel is drilled without the need to accommodate for subsequent femoral tunnel reaming, a more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. Although this MRI analysis of postoperative subjects does indicate improved re-creation of native anatomy by use of the AM technique compared with the TT technique, a difference in functional outcomes favoring one technique over another with respect to graft function and patient outcomes remains undetermined. We acknowledge that coronal plane obliquity was actually increased relative to the native ACL using the AM technique compared with the TT technique. It would appear that when using this single-bundle construct, a tradeoff occurs in optimizing sagittal versus coronal obliquity. Increased coronal obliquity may contribute to constraint of tibial internal rotation. In a cross-sectional study examining gait analysis after ACL reconstruction, Scanlan et al.39 found a negative correlation between the peak external knee flexion moment during walking and the coronal angle of the ACL graft; they posited that with more vertical coronal orientation, patients reduce their net quadriceps usage during walking. The clinical implications of this finding, and of the findings reported in our study, remain unknown. A weakness of this study is the inability to assess
graft position relative to individual AM and PL ACL bundles. Recent biomechanical analyses have shown that vertical “graft mismatch” connecting the AM bundle femoral insertion with the PL bundle tibial insertion has inferior kinematic behavior relative to matched reconstructions, whereas a horizontal “mismatch” connecting the PL femur with the AM tibia succeeds in limiting translation with both Lachman and pivot-shift maneuvers.20 We suspect that the configuration achieved with our AM portal ACL reconstructions mimics this PL femur–AM tibia construct, particularly given the more horizontal position achieved in the coronal plane. However, because of limitations of our MRI reconstructions, we were unable to reliably differentiate between the 2 bundles consistently throughout the length of the contralateral, native ACL. Multiple surgeons were used in this study. However, we observed that tunnel positions were consistent among surgeons using the same technique, and we believe that inclusion of multiple surgeons may serve to reduce the likelihood of bias resulting from the techniques of a single surgeon. In each case the surgeon selected the technique that he or she felt, in his or her hands, would best position the graft based on the patient’s anatomy. This should make findings more ubiquitous to the general orthopaedic community. Nonetheless, we should caution that all reconstructions evaluated were performed by high-volume, fellowship-trained sports medicine surgeons. Furthermore, AM reconstructions are performed at our institution with flexible reaming systems, and our findings may not be generalizable to AM reconstructions with standard, nonflexible reaming systems. Finally, although clear differences between techniques have been elucidated in this study by MRI, no subjective or objective clinical outcomes have been evaluated, because they are beyond the scope of this initial investigation. The true clinical relevance of these MRI measurements remains unknown and is a source of critical ongoing evaluation. CONCLUSIONS In this series AM portal femoral tunnel reaming more accurately restored native ACL anatomy than did TT reaming. Although both techniques can capture the native femoral footprint with similar accuracy, the TT technique requires significantly greater posterior placement of the tibial tunnel, resulting in decreased sagittal graft obliquity. When a tibial tunnel is drilled without the need to accommodate subse-
ACL TUNNEL POSITION AND GRAFT OBLIQUITY quent femoral tunnel reaming, more accurate tibial tunnel position and resultant sagittal graft obliquity are achieved. 16.
Acknowledgment: The authors thank the following physicians for their assistance: Jo A. Hannafin, M.D., Ph.D., Stephen J. O’Brien, M.D., M.B.A., Howard Anthony Rose, M.D., Anne Kelly, M.D., Beth Shubin Stein, M.D., Struan Coleman, M.D., Ph.D., Matthew Koff, Ph.D., and Shirley Magabo, M.D.
17.
18.
REFERENCES 1. Chhabra A, Diduch DR, Blessey PB, Miller MD. Recreating an acceptable angle of the tibial tunnel in the coronal plane in anterior cruciate ligament reconstruction using external landmarks. Arthroscopy 2004;20:328-330. 2. Howell SM, Gittins ME, Gottlieb JE, Traina SM, Zoellner TM. The relationship between the angle of the tibial tunnel in the coronal plane and loss of flexion and anterior laxity after anterior cruciate ligament reconstruction. Am J Sports Med 2001;29:567-574. 3. Dargel J, Schmidt-Wiethoff R, Fischer S, Mader K, Koebke J, Schneider T. Femoral bone tunnel placement using the transtibial tunnel or the anteromedial portal in ACL reconstruction: A radiographic evaluation. Knee Surg Sports Traumatol Arthrosc 2009;17:220-227. 4. Giron F, Cuomo P, Edwards A, Bull AM, Amis AA, Aglietti P. Double-bundle “anatomic” anterior cruciate ligament reconstruction: A cadaveric study of tunnel positioning with a transtibial technique. Arthroscopy 2007;23:7-13. 5. Heming JF, Rand J, Steiner ME. Anatomical limitations of transtibial drilling in anterior cruciate ligament reconstruction. Am J Sports Med 2007;35:1708-1715. 6. Steiner ME, Battaglia TC, Heming JF, Rand JD, Festa A, Baria M. Independent drilling outperforms conventional transtibial drilling in anterior cruciate ligament reconstruction. Am J Sports Med 2009;37:1912-1919. 7. Bedi A, Musahl V, Steuber V, et al. Transtibial versus anteromedial portal reaming in anterior cruciate ligament reconstruction: An anatomic and biomechanical evaluation of surgical technique. Arthroscopy 2011;27:380-390. 8. Bottoni CR. Anterior cruciate ligament femoral tunnel creation by use of anteromedial portal. Arthroscopy 2008;24:1319. 9. Cain EL Jr, Clancy WG Jr. Anatomic endoscopic anterior cruciate ligament reconstruction with patella tendon autograft. Orthop Clin North Am 2002;33:717-725. 10. Harner CD, Honkamp NJ, Ranawat AS. Anteromedial portal technique for creating the anterior cruciate ligament femoral tunnel. Arthroscopy 2008;24:113-115. 11. Lubowitz JH. Anteromedial portal technique for the anterior cruciate ligament femoral socket: Pitfalls and solutions. Arthroscopy 2009;25:95-101. 12. Shelbourne KD, Patel DV. ACL reconstruction using the autogenous bone-patellar tendon-bone graft: Open two-incision technique. Instr Course Lect 1996;45:245-252. 13. Steiner ME, Murray MM, Rodeo SA. Strategies to improve anterior cruciate ligament healing and graft placement. Am J Sports Med 2008;36:176-189. 14. Abebe ES, Moorman CT III, Dziedzic TS, et al. Femoral tunnel placement during anterior cruciate ligament reconstruction: An in vivo imaging analysis comparing transtibial and 2-incision tibial tunnel-independent techniques. Am J Sports Med 2009; 37:1904-1911. 15. Brophy RH, Voos JE, Shannon FJ, et al. Changes in the length
19.
20.
21.
22.
23. 24. 25.
26.
27.
28. 29. 30.
31.
1521
of virtual anterior cruciate ligament fibers during stability testing: A comparison of conventional single-bundle reconstruction and native anterior cruciate ligament. Am J Sports Med 2008;36:2196-2203. Forsythe B, Kopf S, Wong AK, et al. The location of femoral and tibial tunnels in anatomic double-bundle anterior cruciate ligament reconstruction analyzed by three-dimensional computed tomography models. J Bone Joint Surg Am 2010;92: 1418-1426. Gavriilidis I, Motsis EK, Pakos EE, Georgoulis AD, Mitsionis G, Xenakis TA. Transtibial versus anteromedial portal of the femoral tunnel in ACL reconstruction: A cadaveric study. Knee 2008;15:364-367. Kopf S, Forsythe B, Wong AK, et al. Nonanatomic tunnel position in traditional transtibial single-bundle anterior cruciate ligament reconstruction evaluated by three-dimensional computed tomography. J Bone Joint Surg Am 2010;92:14271431. Miller MD, Gerdeman AC, Miller CD, et al. The effects of extra-articular starting point and transtibial femoral drilling on the intra-articular aperture of the tibial tunnel in ACL reconstruction. Am J Sports Med 2010;38:707-712. Pearle AD, Shannon FJ, Granchi C, Wickiewicz TL, Warren RF. Comparison of 3-dimensional obliquity and anisometric characteristics of anterior cruciate ligament graft positions using surgical navigation. Am J Sports Med 2008;36:15341541. Alentorn-Geli E, Lajara F, Samitier G, Cugat R. The transtibial versus the anteromedial portal technique in the arthroscopic bone-patellar tendon-bone anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc 2010;18:1013-1037. Hantes ME, Zachos VC, Liantsis A, Venouziou A, Karantanas AH, Malizos KN. Differences in graft orientation using the transtibial and anteromedial portal technique in anterior cruciate ligament reconstruction: A magnetic resonance imaging study. Knee Surg Sports Traumatol Arthrosc 2009;17:880-886. Bedi A, Altchek DW. The “footprint” anterior cruciate ligament technique: An anatomic approach to anterior cruciate ligament reconstruction. Arthroscopy 2009;25:1128-1138. Asano T, Akagi M, Tanaka K, Tamura J, Nakamura T. In vivo three-dimensional knee kinematics using a biplanar imagematching technique. Clin Orthop Relat Res 2001:157-166. Loh JC, Fukuda Y, Tsuda E, Steadman RJ, Fu FH, Woo SL. Knee stability and graft function following anterior cruciate ligament reconstruction: Comparison between 11 o’clock and 10 o’clock femoral tunnel placement. 2002 Richard O’Connor Award paper. Arthroscopy 2003;19:297-304. Musahl V, Plakseychuk A, VanScyoc A, et al. Varying femoral tunnels between the anatomical footprint and isometric positions: Effect on kinematics of the anterior cruciate ligament-reconstructed knee. Am J Sports Med 2005;33:712-718. Yamamoto Y, Hsu WH, Woo SL, Van Scyoc AH, Takakura Y, Debski RE. Knee stability and graft function after anterior cruciate ligament reconstruction: A comparison of a lateral and an anatomical femoral tunnel placement. Am J Sports Med 2004;32:1825-1832. Lee MC, Seong SC, Lee S, et al. Vertical femoral tunnel placement results in rotational knee laxity after anterior cruciate ligament reconstruction. Arthroscopy 2007;23:771-778. Morgan CD, Kalman KR, Grawl DM. Definitive landmarks for reproducible tibial tunnel placement in anterior cruciate ligament reconstruction. Arthroscopy 1995;11:275-288. Piasecki DP, Bach BR Jr, Espinoza Orias AA, Verma NN. Anterior cruciate ligament reconstruction: Can anatomic femoral placement be achieved with a transtibial technique? Am J Sports Med 2011;39:1306-1315. Arnold MP, Kooloos J, van Kampen A. Single-incision tech-
1522
32.
33.
34. 35.
A. L. BOWERS ET AL.
nique misses the anatomical femoral anterior cruciate ligament insertion: A cadaver study. Knee Surg Sports Traumatol Arthrosc 2001;9:194-199. Ahn JH, Lee SH, Yoo JC, Ha HC. Measurement of the graft angles for the anterior cruciate ligament reconstruction with transtibial technique using postoperative magnetic resonance imaging in comparative study. Knee Surg Sports Traumatol Arthrosc 2007;15:1293-1300. Stanford FC, Kendoff D, Warren RF, Pearle AD. Native anterior cruciate ligament obliquity versus anterior cruciate ligament graft obliquity: An observational study using navigated measurements. Am J Sports Med 2009;37:114-119. Giron F, Buzzi R, Aglietti P. Femoral tunnel position in anterior cruciate ligament reconstruction using three techniques. A cadaver study. Arthroscopy 1999;15:750-756. Albuquerque RF, Amatuzzi MM, Pacheco AP, Angelini FJ, Campos O Jr. Positioning of the femoral tunnel for arthroscopic reconstruction of the anterior cruciate ligament: Comparative study of 2 techniques. Clinics (Sao Paulo) 2007;62:613-618.
36. Purnell ML, Larson AI, Clancy W. Anterior cruciate ligament insertions on the tibia and femur and their relationships to critical bony landmarks using high-resolution volume-rendering computed tomography. Am J Sports Med 2008;36:20832090. 37. Voos JE, Musahl V, Maak TG, Wickiewicz TL, Pearle AD. Comparison of tunnel positions in single-bundle anterior cruciate ligament reconstructions using computer navigation. Knee Surg Sports Traumatol Arthrosc 2010;18:12821289. 38. Iriuchishima T, Tajima G, Ingham SJ, Shen W, Smolinski P, Fu FH. Impingement pressure in the anatomical and nonanatomical anterior cruciate ligament reconstruction: A cadaver study. Am J Sports Med 2010;38:1611-1617. 39. Scanlan SF, Blazek K, Chaudhari AM, Safran MR, Andriacchi TP. Graft orientation influences the knee flexion moment during walking in patients with anterior cruciate ligament reconstruction. Am J Sports Med 2009;37:2173-2178.