- Email: [email protected]
Magnetic Resonance Imaging of 3-Dimensional In Vivo Tibiofemoral Kinematics in Anterior Cruciate Ligament–Reconstructed Knees
Magnetic Resonance Imaging of 3-Dimensional In Vivo Tibiofemoral Kinematics in Anterior Cruciate Ligament–Reconstructed Knees R. Dana Carpenter, Ph.D., Sharmila Majumdar, Ph.D., and C. Benjamin Ma, M.D.
Purpose: The purpose of this study was to use magnetic resonance imaging (MRI) to determine 3-dimensional knee kinematics after anterior cruciate ligament (ACL) reconstruction. Methods: Nine ACL-reconstructed and contralateral knees were tested 12 ⫾ 8 months after surgery. MRI was performed at full extension and 40° of knee flexion under simulated weight-bearing conditions. Femoral condyle positions, tibial rotation, contact area, and contact location were analyzed by use of MRI-based 3-dimensional models. Results: When knees were fully extended, tibiae in ACLreconstructed knees were externally rotated by 3.6° ⫾ 4.2° compared with contralateral knees. The external rotation was due to anterior subluxation of the medial side of the tibia. At 40° of knee flexion, tibiae in ACL-reconstructed knees and contralateral knees were both internally rotated by 5.3°. There were no significant differences in contact area or contact location between ACLreconstructed and contralateral knees. When moving from extension to flexion, ACL-reconstructed knees exhibited 3.5° ⫾ 5.9° more internal tibial rotation than contralateral knees. Conclusions: Reconstruction of the ACL restored normal motion on the lateral side of the knee but not on the medial side, resulting in increased internal tibial rotation when moving from full extension to 40° of flexion. These results suggest that ACL reconstruction does not restore normal kinematics on the medial side of the knee, which may lead to early cartilage degeneration. Level of Evidence: Level IV, therapeutic case series. Key Words: Anterior cruciate ligament reconstruction—Magnetic resonance imaging—Kinematics—Biomechanics—Internal tibial rotation—Arthrosis.
A
ltered mechanical loading due to kinematic changes in the knee is thought to be an important factor in cartilage degeneration and progression of osteoarthritis (OA).1,2 Rupture of the anterior cruciate ligament (ACL) causes changes in tibiofemoral kinematics that may affect the mechanical environment of
From the Departments of Radiology (R.D.C., S.M.) and Orthopaedic Surgery (C.B.M.), University of California, San Francisco, San Francisco, California, U.S.A. Supported by the Orthopaedic Research and Education Foundation and a Young Investigator Grant from the American Orthopaedic Society for Sports Medicine. The authors report no conflict of interest. Received August 29, 2008; accepted January 21, 2009. Address correspondence and reprint requests to R. Dana Carpenter, Ph.D., Department of Radiology, University of California, San Francisco, 185 Berry St, Suite 350, San Francisco, CA 94143, U.S.A. E-mail: [email protected] © 2009 by the Arthroscopy Association of North America 0749-8063/09/2507-8501$36.00/0 doi:10.1016/j.arthro.2009.01.014
760
the articular cartilage and subchondral bone, making ACL injury a potential risk factor for OA.3-6 Reconstruction of the ACL diminishes the severity of postreconstruction OA by decreasing the number of meniscal tears, which predispose patients for OA.7,8 However, similar rates of OA have been observed in patients who undergo reconstruction and patients who decide not to undergo surgery. A long-term study of young, female soccer players 12 years after ACL injury showed that in those who underwent ACL reconstruction, symptomatic knee OA developed at a rate similar to that in those who chose not to have ACL reconstruction.9 In another study 84% of patients after ACL reconstruction had slight to moderate changes equivalent to OA after 20 years.10 Short-term changes have also been observed. Arthroscopic examination of 105 patients who had ACL reconstruction 15 months after surgery showed degeneration of all cartilage surfaces except the lateral femoral condyle.11
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 25, No 7 (July), 2009: pp 760-766
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION Several different image-based techniques have been developed to measure in vivo kinematics of ACLdeficient and ACL-reconstructed knees. In vivo magnetic resonance imaging (MRI) during full or partial weight bearing has been used in a number of studies to measure tibiofemoral translation and contact location based on a pair of sagittal image slices bisecting the medial and lateral femoral condyles.12-17 Logan et al.13,14 used this technique to find that the position of the lateral tibial plateau in both ACL-deficient and ACL-reconstructed knees was anterior to that in normal contralateral knees. Anterior subluxation of the lateral tibia in ACL-reconstructed knees existed despite a return to normal anteroposterior laxity. Previous studies have developed a 3-dimensional (3D) analysis of in vivo tibiofemoral kinematics, contact area, and contact location with an MRI-compatible loading device to simulate partial weight bearing during imaging in a clinical MRI system.4,18 Using this technique, a study of 8 patients with unilateral ACL injury showed that the position of the lateral tibia in ACL-deficient knees in full extension was anterior to that in normal knees.4 The results also showed that the cartilage-on-cartilage contact location on the tibial plateau in ACL-deficient knees was posterior to that in normal knees and that the contact centroid on the medial tibial plateau translated less than in normal knees during movement from full extension to 45° of flexion. Knee kinematics and cartilage contact characteristics affect the mechanical environment in both cartilage and bone. Therefore the altered kinematics found in previous studies of ACL-deficient knees may have important effects on joint health. Measuring the 3D kinematics of ACL-reconstructed knees provides both a quantitative evaluation of knee performance after surgery and a means of understanding how the biomechanical environment in the knee changes after reconstruction. The goal of this study was to determine 3D flexion/ extension knee kinematics in ACL-reconstructed knees. Our hypothesis was that ACL-reconstructed knees do not have normal knee kinematics when compared with contralateral normal knees. METHODS We recruited 9 patients (5 men and 4 women; mean age, 32 ⫾ 9 years) with unilateral ACL reconstruction for the study. One patient received an Achilles tendon allograft, two received bone-to-bone autografts, and the rest received hamstring tendon autografts. A singleincision transtibial technique was used by the same sur-
761
geon to perform all ACL reconstructions. For the tibial tunnel position, all tibial tunnels were placed through the remaining tibial stump with the posterior edge of the anterior horn of the lateral meniscus as the landmark. The tibial stump was left behind to enable accurate placement of the graft. The femoral tunnel was positioned by use of the transtibial technique. A 6-mm offset guide was used in all cases, and the tunnels were placed at the 1:30 and 10:30 positions around the notch. Imaging was performed at a minimum of 6 months after surgery, with a mean time of 12 ⫾ 8 months (range, 6 to 30 months) postoperatively. All patients had a clinically normal Lachman examination and instrumented examinations with anteroposterior translation of ⫺1 to 2 mm when compared with the contralateral uninjured knees. None of the patients had meniscectomy, but 3 patients had all-inside lateral meniscus repairs. There were no patients in this study who had cartilage abnormalities that required further treatment. All patients gave informed consent to participate in the study, and all procedures were approved by the Committee on Human Research at our institution. In vivo imaging under simulated partial weightbearing conditions and 3D kinematic analysis were performed with methods previously used in a study of ACL-injured patients.4 Patients were imaged in a General Electric 3.0-T Signa MRI system (GE Healthcare, Waukesha, WI). A dual-element, phased-array paddle coil (Nova Medical, Wilmington, MA) was attached to the medial and lateral sides of the knee, and an MRI-compatible loading device was used to apply a compressive force of 125 N to the bottom of the patient’s foot while the patient lay in a supine position (Fig 1). Proton density–weighted images of the knee were acquired with a fast spin echo pulse sequence in the sagittal plane with a field of view of 16 cm, 512 ⫻ 256 matrix, in-plane resolution of 0.3 mm, slice thickness of 1.5 mm, repetition time of 3,500 milliseconds, and echo time of 9.7 milliseconds. Sixty-six slices were obtained with an imaging time of 7 minutes 40 seconds. Images of each knee were obtained in a position of full extension and in a position of approximately 40° of flexion. To help provide a consistent flexion position, patients were asked to flex the knee until the patella rested against a positioning plate attached to the scanning table (Fig 1). For analysis of knee kinematics, the femur and tibia of each image were semiautomatically segmented by use of B-splines created with in-house software written and run in MATLAB (The MathWorks, Natick, MA). Medial and lateral tibiofemoral cartilage-on-
762
R. D. CARPENTER ET AL.
FIGURE 1. Experimental setup. Weights are hung behind the patient in the MRI system, and a set of pulleys and a loading plate transfer the force into a compressive load at the foot. The phasedarray paddle coil is attached to the medial and lateral sides of the knee, and a knee-positioning plate provides feedback to help ensure a consistent angle of knee flexion.
cartilage contact regions were also defined by use of B-splines. The shape of the tibia in the flexed position was registered to that in the fully extended position by use of an iterative closest-point shape-matching algorithm.19 Thus the tibia was held fixed, and kinematic parameters were measured by analyzing the motion of the femur relative to the tibia (Fig 2). The medial-lateral (M-L) axis of the tibia was automatically defined as the line connecting the most posterior points of the medial and lateral sides of the tibial plateau, and the midpoint of this line defined the origin of the tibial coordinate system. The tibial shaft
FIGURE 2. Image analysis process. Sagittal images of the tibia and femur in an extended position (red) and flexed position (green) were segmented (A), and all slices were combined to obtain 3D shapes (B). The shapes of the tibiae were matched, and the motion of the femur relative to the tibia was analyzed (C).
FIGURE 3. Coordinate systems used to analyze kinematics of femur (left) and tibia (right). The centers of spheres fit to the femoral condyles were used to quantify femoral condyle locations, and the most posterior points on the medial and lateral sides of the tibial plateau were used to define the tibial coordinate system.
axis was manually defined on a central sagittal slice, and the anterior-posterior (A-P) axis of the tibia was defined by taking the cross product between the M-L axis and the tibial shaft axis. The inferior-superior axis of the tibia was then defined by taking the cross product between the M-L and A-P axes, providing a mutually orthogonal set of anatomic axes for the tibia (Fig 3). To establish an anatomic coordinate system for the femur, spheres were automatically fit to the contours defining the posterior surfaces of the femoral condyles by use of a least-squares fitting algorithm. The line joining the 2 sphere centers defined the M-L axis, and the midpoint of this line served as the origin for the femoral coordinate system. The femoral shaft axis was manually defined on a central sagittal slice, and the A-P axis of the femur was defined by taking the cross product between the M-L axis and the femoral shaft axis. The inferior-superior axis of the femur was then defined by taking the cross product between the M-L and A-P axes, providing a mutually orthogonal set of anatomic axes for the femur (Fig 2). Positions of the femoral condyle centers measured in flexion and extension were expressed relative to the tibial origin. The angle of tibial rotation at each position was measured as the angle between the M-L axes of the femur and tibia projected onto the transverse plane of the tibia. The position of the femoral origin relative to the tibial origin was also recorded. Translations of the condyle centers, translation of the femoral origin, and amount of tibial rotation when moving from the extended position to the flexed position were computed by taking the differences between values in the flexed and extended positions. Measurements with a rotational phantom indicate an accuracy of 0.1° for tibiofemoral rotations. The
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION TABLE 1.
Internal Tibial Rotation in 9 Patients After ACL Reconstruction
Knee Position
ITR in ACL-Reconstructed Knees
ITR in Contralateral Knees
Extension 40° of flexion Extension to 40° of flexion
⫺4.3° ⫾ 4.7°* 5.3° ⫾ 3.7° 9.5° ⫾ 5.3°*
0.7° ⫾ 5.0°* 5.3° ⫾ 3.9° 6.0° ⫾ 6.0°*
NOTE. Data are given as mean ⫾ SD. Positions in full extension and approximately 40° of flexion are provided along with the amount of rotation that occurred when moving from the extended position to the flexed position. A positive value indicates internal tibial rotation, and a negative value indicates external tibial rotation. Abbreviation: ITR, internal tibial rotation. *Significant difference between ACL-reconstructed and contralateral knees (P ⬍ .05).
intraobserver and interobserver reproducibilities of the analysis techniques used in this study amount to 2 to 3 pixels (0.6 to 0.9 mm) for translations and approximately 1.5° for rotation.4 Cartilage-on-cartilage contact areas in the medial and lateral compartments were computed by connecting all spline points with a set of triangles and then summing the triangle areas. The contact centroid in each compartment was defined as the area centroid for the corresponding set of triangles. This method provided the overall center of contact for the compartment as a whole. Contact centroid coordinates were expressed with respect to the tibial origin, and centroid translations when moving from the extended position to the flexed position were computed as the difference between the locations in the 2 positions. For each measured parameter, a paired Student t test was performed to test for a significant difference (␣ ⫽ .05) between reconstructed knees and contralateral knees. Paired Student t tests were also used to detect significant changes in contact area and centroid location when moving from the extended position to the flexed position. RESULTS In the fully extended position (6.3° of hyperextension), the tibia was externally rotated by a mean of 4.3° in ACL-reconstructed knees, whereas the tibia in the contralateral limbs in the fully extended position (5.5° of hyperextension) was not significantly rotated relative to the femur (Table 1). In the flexed position, tibiae in ACL-reconstructed knees and contralateral knees were both internally rotated by a mean of 5.3°.
763
Because of the externally rotated position of the tibia in full extension, the tibia in ACL-reconstructed knees underwent a significantly larger amount of internal tibial rotation when moving from the extended position to the flexed position. Cartilage-on-cartilage contact area in ACL-reconstructed knees and contralateral knees decreased significantly in the medial compartment when moving from extension to flexion (Table 2). A decrease in contact area occurred in the lateral compartment in contralateral knees, but the contact area in the lateral compartment in ACL-reconstructed knees did not change significantly. When moving from extension to flexion, the contact centroid in the medial compartment underwent significant posterior translation in ACL-reconstructed and contralateral knees. The contact centroid in the lateral compartment underwent significant posterior and lateral translation in contralateral knees but not in ACL-reconstructed knees. No significant differences in contact area in either position or contact centroid translation were detected between ACL-reconstructed and contralateral knees. In the fully extended position, the A-P alignment of the whole femur in ACL-reconstructed knees was not significantly different than that in contralateral knees (Table 3). However, when the positions of the medial and lateral condyles were analyzed separately, the A-P
TABLE 2. Cartilage-on-Cartilage Contact Centroid Translations and Contact Areas in Patients After ACL Reconstruction ACL-Reconstructed Knees Translation of medial centroid (mm) A-P M-L Translation of lateral centroid (mm) A-P M-L Medial contact area (mm2) Extension Flexion Lateral contact area (mm2) Extension Flexion
5.5 ⫾ 3.0* 0.9 ⫾ 2.1
Contralateral Knees
6.2 ⫾ 2.1* 0.8 ⫾ 2.4
3.0 ⫾ 3.9 ⫺0.3 ⫾ 1.4
2.2 ⫾ 2.7* ⫺1.4 ⫾ 1.5*
308 ⫾ 77* 207 ⫾ 63
298 ⫾ 63* 189 ⫾ 40
174 ⫾ 52 159 ⫾ 69
195 ⫾ 45* 143 ⫾ 62
NOTE. Translations are expressed in the coordinate system of the tibia. Positive values indicate posterior translation and medial translation. *Significant change when moving from extension to flexion (P ⬍ .05).
764
R. D. CARPENTER ET AL. TABLE 3.
Side-to-Side Differences in Alignment of Femoral Condyles and Whole Femur in Fully Extended Position
Position in Full Extension A-P alignment of femur (mm) Internal tibial rotation (°) A-P condyle alignment (mm) Medial condyle Lateral condyle
ACL Reconstruction Patients (Reconstructed ⫺ Contralateral) (n ⫽ 9)
ACL-Deficient Patients (Injured ⫺ Contralateral) (n ⫽ 8)
1.0 ⫾ 2.3 ⫺3.6 ⫾ 4.2*†
2.6 ⫾ 1.7* 4.9 ⫾ 5.5*†
2.6 ⫾ 2.2* ⫺0.2 ⫾ 4.0†
1.4 ⫾ 2.9 3.8 ⫾ 2.0*†
NOTE. Data are given as mean ⫾ SD. A positive value indicates a more posterior alignment, a more medial alignment, or a more internally rotated position. Data for ACL-deficient knees are from Shefelbine et al.4 *Significant difference between affected and contralateral knees (P ⬍ .05). †Significant difference between ACL-deficient and ACL reconstruction patients (P ⬍ .05).
position of the medial femoral condyle in ACL-reconstructed knees in full extension was significantly (P ⬍ .05) posterior to that in contralateral knees whereas no difference was present on the lateral side. The relative positions of the medial and lateral condyles produced the externally rotated tibia position observed in the fully extended position. DISCUSSION The results of this study suggest that ACL reconstruction partially restores normal knee kinematics, but some important differences in kinematics between ACL-reconstructed and contralateral knees were observed. The kinematics of the lateral compartment were restored between extension and flexion, whereas those in the medial compartment were not, resulting in increasing internal tibial rotation between extension and flexion in the ACL-reconstructed knees. We compared our results with those in a previous cohort of ACL-deficient patients tested under similar loading conditions.4 Reconstructed knees were imaged at flexion angles of ⫺5.9° ⫾ 4.4° in the fully extended position and 39.7° ⫾ 12.1° in the flexed position. For comparison, the previously analyzed ACL-deficient patients were imaged at ⫺9.9° ⫾ 4.7° in the fully extended position and 42.7° ⫾ 8.3° in the flexed position. Because the mean angles in the 2 positions were not significantly different (P ⫽ .17 in extension and P ⫽ .60 in flexion) between the 2 study populations, comparisons between the 2 groups of patients were deemed reasonable. In the fully extended position the posterior alignment of the lateral condyle observed in ACL-deficient knees was restored to a normal position in ACLreconstructed knees, and normal A-P alignment of the
whole femur was also restored (Table 3). However, the center of the medial femoral condyle in ACLreconstructed knees was 2.6 mm posterior to that in contralateral knees in full extension (i.e., the medial side of the tibia was anteriorly subluxated). When ACL-reconstructed knees moved from the abnormal, externally rotated position in extension to a normally rotated position in flexion, ACL-reconstructed knees exhibited a total of 3.5° more internal tibial rotation than contralateral knees (Table 1). This result is similar to that obtained by Georgoulis et al.,20 who found that the tibia in ACL-reconstructed knees underwent approximately 4° more total rotation than normal knees during the stance phase of gait. Although the mechanical loads at the knee are very different between gait studies and our loading device, the 2 positions used in our study may capture a range of knee motion similar to that which occurs during the stance phase of gait. Our measurements of internal/external tibial rotation also correspond well with an in vivo study of running performed by Tashman et al.21 In that study ACL-reconstructed knees were slightly externally rotated (approximately 2° to 3°) at foot strike, which was also near the moment of maximum knee extension. Normal knees, on the other hand, were internally rotated by approximately 1° at foot strike. Then, as the knee flexed during the stance phase, the tibia rotated internally in both types of knees, reaching an internal rotation of 5° to 10°. Although the loading configuration and dynamic environment during running are very different from those of the simplified extension/ flexion configuration used in our study, our results (Table 1) show a range of tibial internal/external rotation and a difference between ACL-reconstructed
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION and intact knees consistent with the results of Tashman et al. Our image-based method allowed analysis of the relative contributions of medial and lateral knee motion to knee rotational kinematics. Our results suggest that ACL reconstruction restores a normal amount of A-P motion on the lateral side of the knee but not on the medial side of the knee. This finding may have implications in understanding the effects of ACL injury and reconstruction on OA progression, which occurs predominantly in the medial compartment. Despite the difference in rotational kinematics, cartilage-on-cartilage contact areas and locations were not significantly different between ACL-reconstructed and contralateral knees, which is encouraging in terms of restoring a normal biomechanical environment to the ACL-injured knee. In the previous study of ACLdeficient knees performed in our laboratory, the locations of the medial and lateral contact centroids were posterior to those in contralateral knees.4 It appears that ACL reconstruction in our cohort restored contact centroids to their normal locations. The sizes of the contact regions were also not significantly different between ACL-reconstructed and contralateral knees, suggesting that the joint contact force is distributed over a similar area of cartilage in reconstructed knees. However, although the contact area and centroids may be close to normal, the abnormal initial tibial rotation and the resulting increase in internal tibial rotation when moving from extension to flexion may affect the distribution of loads in the ACL-reconstructed knee. For example, a study using finite-element models showed that 5° of increased internal tibial rotation after ACL injury leads to changes in cartilage stresses that may subsequently accelerate the rate of cartilage thinning.1 The loading configuration used in this study was intended to simulate partial weight bearing. Other studies have used dual-plane fluoroscopy22 and openMRI systems14 to analyze ACL-reconstructed kinematics during weight bearing. In patients in full knee extension during weight bearing, Papannagari et al.22 found that the tibia in the ACL-reconstructed knee was shifted by a mean of 2.9 mm anterior to that in contralateral knees. Although we did not observe this anterior shift for the whole tibia, we did see a similar shift (2.6 mm) on the medial side only. This finding again illustrates the value of analyzing the medial and lateral compartments separately. Also unlike Papannagari et al., we did not observe significant anterior translation of the tibia during flexion. It is likely that this difference is related to our use of a femur coor-
765
dinate system based on the centers of spheres fit to the femoral condyles instead of the transepicondylar axis. Logan et al.14 used images acquired with an open-MRI system to analyze medial and lateral femoral condyle motion during knee flexion using a single sagittal slice through each condyle. They found that the lateral side of the tibia in ACL-reconstructed knees was always shifted 5 mm in the anterior direction at each of 4 flexion positions from 0° to 90°. We did not find any significant differences in lateral femoral condyle motion between ACL-reconstructed and contralateral knees. This discrepancy could be because we used the positions of 3D sphere centers instead of circles on single sagittal slices to track condyle positions, or it may be a result of different surgical techniques. This study had some limitations that should be considered when interpreting the results. First, the loading configuration was not the same as that which occurs in full weight bearing or during functional activities such as walking. However, as stated previously, the 125-N compressive load used in this study was sufficient to produce a result (anterior subluxation of the tibia in extension) similar to that seen in full weight bearing with dual-plane fluoroscopy. In addition, although the amount of knee flexion observed in our study was similar to that which occurs during the stance phase of gait, space limitations prevented us from examining the knee at higher degrees of flexion, such as those that occur during squatting or step-up activities.23 It may be important for future studies to analyze kinematics at a higher angle of flexion than that used in our study. The use of 2 static knee positions to compute translations and rotations may also miss differences in knee motion that occur between these 2 positions. Finally, the ACL reconstructions were performed with a variety of graft and fixation choices, and it would be ideal to examine only a single graft type. However, the ACL reconstructions were performed by the same surgeon using the same femoral and tibial tunnel placements, minimizing surgeon and tunnel position variability. Future studies will include a larger sample size and investigate the effect of various ACL reconstructions such as doublebundle ACL reconstruction and medial portal drilling of femoral tunnels. CONCLUSIONS Reconstruction of the ACL restored normal motion on the lateral side of the knee but not on the medial side, resulting in increased internal tibial rotation when moving from full extension to 40° of flexion.
766
R. D. CARPENTER ET AL.
These results suggest that ACL reconstruction does not restore normal kinematics on the medial side of the knee, which may lead to early cartilage degeneration. Acknowledgment: The authors thank Sandra Shefelbine for the use of her data from ACL-deficient knees.
REFERENCES 1. Andriacchi TP, Briant PL, Bevill SL, Koo S. Rotational changes at the knee after ACL injury cause cartilage thinning. Clin Orthop Relat Res 2006;442:39-44. 2. Andriacchi TP, Mundermann A. The role of ambulatory mechanics in the initiation and progression of knee osteoarthritis. Curr Opin Rheumatol 2006;18:514-518. 3. Defrate LE, Papannagari R, Gill TJ, Moses JM, Pathare NP, Li G. The 6 degrees of freedom kinematics of the knee after anterior cruciate ligament deficiency: An in vivo imaging analysis. Am J Sports Med 2006;34:1240-1246. 4. Shefelbine SJ, Ma CB, Lee KY, et al. MRI analysis of in vivo meniscal and tibiofemoral kinematics in ACL-deficient and normal knees. J Orthop Res 2006;24:1208-1217. 5. Li G, Papannagari R, DeFrate LE, Yoo JD, Park SE, Gill TJ. The effects of ACL deficiency on mediolateral translation and varus-valgus rotation. Acta Orthop 2007;78:355-360. 6. Andriacchi TP, Mundermann A, Smith RL, Alexander EJ, Dyrby CO, Koo S. A framework for the in vivo pathomechanics of osteoarthritis at the knee. Ann Biomed Eng 2004;32:447457. 7. Church S, Keating JF. Reconstruction of the anterior cruciate ligament: Timing of surgery and the incidence of meniscal tears and degenerative change. J Bone Joint Surg Br 2005;87: 1639-1642. 8. Ohly NE, Keating JF. Arthrosis following ACL reconstruction. In: Prodromos C, Brown C, Fu FH, Georgoulis AD, eds. The anterior cruciate ligament: Reconstruction and basic science. Philadelphia: Saunders, 2008;69-79. 9. Lohmander LS, Ostenberg A, Englund M, Roos H. High prevalence of knee osteoarthritis, pain, and functional limitations in female soccer players twelve years after anterior cruciate ligament injury. Arthritis Rheum 2004;50:3145-3152. 10. Maletius W, Messner K. Eighteen- to twenty-four-year follow-up after complete rupture of the anterior cruciate ligament. Am J Sports Med 1999;27:711-717.
11. Asano H, Muneta T, Ikeda H, Yagishita K, Kurihara Y, Sekiya I. Arthroscopic evaluation of the articular cartilage after anterior cruciate ligament reconstruction: A short-term prospective study of 105 patients. Arthroscopy 2004;20:474-481. 12. Johal P, Williams A, Wragg P, Hunt D, Gedroyc W. Tibiofemoral movement in the living knee. A study of weight bearing and non-weight bearing knee kinematics using ‘interventional’ MRI. J Biomech 2005;38:269-276. 13. Logan M, Dunstan E, Robinson J, Williams A, Gedroyc W, Freeman M. Tibiofemoral kinematics of the anterior cruciate ligament (ACL)-deficient weightbearing, living knee employing vertical access open “interventional” multiple resonance imaging. Am J Sports Med 2004;32:720-726. 14. Logan MC, Williams A, Lavelle J, Gedroyc W, Freeman M. Tibiofemoral kinematics following successful anterior cruciate ligament reconstruction using dynamic multiple resonance imaging. Am J Sports Med 2004;32:984-992. 15. Scarvell JM, Smith PN, Refshauge KM, Galloway HR, Woods KR. Evaluation of a method to map tibiofemoral contact points in the normal knee using MRI. J Orthop Res 2004;22:788-793. 16. Scarvell JM, Smith PN, Refshauge KM, Galloway HR, Woods KR. Association between abnormal kinematics and degenerative change in knees of people with chronic anterior cruciate ligament deficiency: A magnetic resonance imaging study. Aust J Physiother 2005;51:233-240. 17. Scarvell JM, Smith PN, Refshauge KM, Galloway HR. Magnetic resonance imaging analysis of kinematics in osteoarthritic knees. J Arthroplasty 2007;22:383-393. 18. Lee KY, Slavinsky JP, Ries MD, Blumenkrantz G, Majumdar S. Magnetic resonance imaging of in vivo kinematics after total knee arthroplasty. J Magn Reson Imaging 2005;21:172178. 19. Besl PJ, McKay ND. A method for registration of 3-D shapes. IEEE Trans Pattern Anal Mach Intell 1992;14:239-256. 20. Georgoulis AD, Ristanis S, Chouliaras V, Moraiti C, Stergiou N. Tibial rotation is not restored after ACL reconstruction with a hamstring graft. Clin Orthop Relat Res 2007;454:89-94. 21. Tashman S, Collon D, Anderson K, Kolowich P, Anderst W. Abnormal rotational knee motion during running after anterior cruciate ligament reconstruction. Am J Sports Med 2004;32: 975-983. 22. Papannagari R, Gill TJ, Defrate LE, Moses JM, Petruska AJ, Li G. In vivo kinematics of the knee after anterior cruciate ligament reconstruction: A clinical and functional evaluation. Am J Sports Med 2006;34:2006-2012. 23. Andriacchi TP, Dyrby CO. Interactions between kinematics and loading during walking for the normal and ACL deficient knee. J Biomech 2005;38:293-298.
Purpose: The purpose of this study was to use magnetic resonance imaging (MRI) to determine 3-dimensional knee kinematics after anterior cruciate ligament (ACL) reconstruction. Methods: Nine ACL-reconstructed and contralateral knees were tested 12 ⫾ 8 months after surgery. MRI was performed at full extension and 40° of knee flexion under simulated weight-bearing conditions. Femoral condyle positions, tibial rotation, contact area, and contact location were analyzed by use of MRI-based 3-dimensional models. Results: When knees were fully extended, tibiae in ACLreconstructed knees were externally rotated by 3.6° ⫾ 4.2° compared with contralateral knees. The external rotation was due to anterior subluxation of the medial side of the tibia. At 40° of knee flexion, tibiae in ACL-reconstructed knees and contralateral knees were both internally rotated by 5.3°. There were no significant differences in contact area or contact location between ACLreconstructed and contralateral knees. When moving from extension to flexion, ACL-reconstructed knees exhibited 3.5° ⫾ 5.9° more internal tibial rotation than contralateral knees. Conclusions: Reconstruction of the ACL restored normal motion on the lateral side of the knee but not on the medial side, resulting in increased internal tibial rotation when moving from full extension to 40° of flexion. These results suggest that ACL reconstruction does not restore normal kinematics on the medial side of the knee, which may lead to early cartilage degeneration. Level of Evidence: Level IV, therapeutic case series. Key Words: Anterior cruciate ligament reconstruction—Magnetic resonance imaging—Kinematics—Biomechanics—Internal tibial rotation—Arthrosis.
A
ltered mechanical loading due to kinematic changes in the knee is thought to be an important factor in cartilage degeneration and progression of osteoarthritis (OA).1,2 Rupture of the anterior cruciate ligament (ACL) causes changes in tibiofemoral kinematics that may affect the mechanical environment of
From the Departments of Radiology (R.D.C., S.M.) and Orthopaedic Surgery (C.B.M.), University of California, San Francisco, San Francisco, California, U.S.A. Supported by the Orthopaedic Research and Education Foundation and a Young Investigator Grant from the American Orthopaedic Society for Sports Medicine. The authors report no conflict of interest. Received August 29, 2008; accepted January 21, 2009. Address correspondence and reprint requests to R. Dana Carpenter, Ph.D., Department of Radiology, University of California, San Francisco, 185 Berry St, Suite 350, San Francisco, CA 94143, U.S.A. E-mail: [email protected] © 2009 by the Arthroscopy Association of North America 0749-8063/09/2507-8501$36.00/0 doi:10.1016/j.arthro.2009.01.014
760
the articular cartilage and subchondral bone, making ACL injury a potential risk factor for OA.3-6 Reconstruction of the ACL diminishes the severity of postreconstruction OA by decreasing the number of meniscal tears, which predispose patients for OA.7,8 However, similar rates of OA have been observed in patients who undergo reconstruction and patients who decide not to undergo surgery. A long-term study of young, female soccer players 12 years after ACL injury showed that in those who underwent ACL reconstruction, symptomatic knee OA developed at a rate similar to that in those who chose not to have ACL reconstruction.9 In another study 84% of patients after ACL reconstruction had slight to moderate changes equivalent to OA after 20 years.10 Short-term changes have also been observed. Arthroscopic examination of 105 patients who had ACL reconstruction 15 months after surgery showed degeneration of all cartilage surfaces except the lateral femoral condyle.11
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 25, No 7 (July), 2009: pp 760-766
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION Several different image-based techniques have been developed to measure in vivo kinematics of ACLdeficient and ACL-reconstructed knees. In vivo magnetic resonance imaging (MRI) during full or partial weight bearing has been used in a number of studies to measure tibiofemoral translation and contact location based on a pair of sagittal image slices bisecting the medial and lateral femoral condyles.12-17 Logan et al.13,14 used this technique to find that the position of the lateral tibial plateau in both ACL-deficient and ACL-reconstructed knees was anterior to that in normal contralateral knees. Anterior subluxation of the lateral tibia in ACL-reconstructed knees existed despite a return to normal anteroposterior laxity. Previous studies have developed a 3-dimensional (3D) analysis of in vivo tibiofemoral kinematics, contact area, and contact location with an MRI-compatible loading device to simulate partial weight bearing during imaging in a clinical MRI system.4,18 Using this technique, a study of 8 patients with unilateral ACL injury showed that the position of the lateral tibia in ACL-deficient knees in full extension was anterior to that in normal knees.4 The results also showed that the cartilage-on-cartilage contact location on the tibial plateau in ACL-deficient knees was posterior to that in normal knees and that the contact centroid on the medial tibial plateau translated less than in normal knees during movement from full extension to 45° of flexion. Knee kinematics and cartilage contact characteristics affect the mechanical environment in both cartilage and bone. Therefore the altered kinematics found in previous studies of ACL-deficient knees may have important effects on joint health. Measuring the 3D kinematics of ACL-reconstructed knees provides both a quantitative evaluation of knee performance after surgery and a means of understanding how the biomechanical environment in the knee changes after reconstruction. The goal of this study was to determine 3D flexion/ extension knee kinematics in ACL-reconstructed knees. Our hypothesis was that ACL-reconstructed knees do not have normal knee kinematics when compared with contralateral normal knees. METHODS We recruited 9 patients (5 men and 4 women; mean age, 32 ⫾ 9 years) with unilateral ACL reconstruction for the study. One patient received an Achilles tendon allograft, two received bone-to-bone autografts, and the rest received hamstring tendon autografts. A singleincision transtibial technique was used by the same sur-
761
geon to perform all ACL reconstructions. For the tibial tunnel position, all tibial tunnels were placed through the remaining tibial stump with the posterior edge of the anterior horn of the lateral meniscus as the landmark. The tibial stump was left behind to enable accurate placement of the graft. The femoral tunnel was positioned by use of the transtibial technique. A 6-mm offset guide was used in all cases, and the tunnels were placed at the 1:30 and 10:30 positions around the notch. Imaging was performed at a minimum of 6 months after surgery, with a mean time of 12 ⫾ 8 months (range, 6 to 30 months) postoperatively. All patients had a clinically normal Lachman examination and instrumented examinations with anteroposterior translation of ⫺1 to 2 mm when compared with the contralateral uninjured knees. None of the patients had meniscectomy, but 3 patients had all-inside lateral meniscus repairs. There were no patients in this study who had cartilage abnormalities that required further treatment. All patients gave informed consent to participate in the study, and all procedures were approved by the Committee on Human Research at our institution. In vivo imaging under simulated partial weightbearing conditions and 3D kinematic analysis were performed with methods previously used in a study of ACL-injured patients.4 Patients were imaged in a General Electric 3.0-T Signa MRI system (GE Healthcare, Waukesha, WI). A dual-element, phased-array paddle coil (Nova Medical, Wilmington, MA) was attached to the medial and lateral sides of the knee, and an MRI-compatible loading device was used to apply a compressive force of 125 N to the bottom of the patient’s foot while the patient lay in a supine position (Fig 1). Proton density–weighted images of the knee were acquired with a fast spin echo pulse sequence in the sagittal plane with a field of view of 16 cm, 512 ⫻ 256 matrix, in-plane resolution of 0.3 mm, slice thickness of 1.5 mm, repetition time of 3,500 milliseconds, and echo time of 9.7 milliseconds. Sixty-six slices were obtained with an imaging time of 7 minutes 40 seconds. Images of each knee were obtained in a position of full extension and in a position of approximately 40° of flexion. To help provide a consistent flexion position, patients were asked to flex the knee until the patella rested against a positioning plate attached to the scanning table (Fig 1). For analysis of knee kinematics, the femur and tibia of each image were semiautomatically segmented by use of B-splines created with in-house software written and run in MATLAB (The MathWorks, Natick, MA). Medial and lateral tibiofemoral cartilage-on-
762
R. D. CARPENTER ET AL.
FIGURE 1. Experimental setup. Weights are hung behind the patient in the MRI system, and a set of pulleys and a loading plate transfer the force into a compressive load at the foot. The phasedarray paddle coil is attached to the medial and lateral sides of the knee, and a knee-positioning plate provides feedback to help ensure a consistent angle of knee flexion.
cartilage contact regions were also defined by use of B-splines. The shape of the tibia in the flexed position was registered to that in the fully extended position by use of an iterative closest-point shape-matching algorithm.19 Thus the tibia was held fixed, and kinematic parameters were measured by analyzing the motion of the femur relative to the tibia (Fig 2). The medial-lateral (M-L) axis of the tibia was automatically defined as the line connecting the most posterior points of the medial and lateral sides of the tibial plateau, and the midpoint of this line defined the origin of the tibial coordinate system. The tibial shaft
FIGURE 2. Image analysis process. Sagittal images of the tibia and femur in an extended position (red) and flexed position (green) were segmented (A), and all slices were combined to obtain 3D shapes (B). The shapes of the tibiae were matched, and the motion of the femur relative to the tibia was analyzed (C).
FIGURE 3. Coordinate systems used to analyze kinematics of femur (left) and tibia (right). The centers of spheres fit to the femoral condyles were used to quantify femoral condyle locations, and the most posterior points on the medial and lateral sides of the tibial plateau were used to define the tibial coordinate system.
axis was manually defined on a central sagittal slice, and the anterior-posterior (A-P) axis of the tibia was defined by taking the cross product between the M-L axis and the tibial shaft axis. The inferior-superior axis of the tibia was then defined by taking the cross product between the M-L and A-P axes, providing a mutually orthogonal set of anatomic axes for the tibia (Fig 3). To establish an anatomic coordinate system for the femur, spheres were automatically fit to the contours defining the posterior surfaces of the femoral condyles by use of a least-squares fitting algorithm. The line joining the 2 sphere centers defined the M-L axis, and the midpoint of this line served as the origin for the femoral coordinate system. The femoral shaft axis was manually defined on a central sagittal slice, and the A-P axis of the femur was defined by taking the cross product between the M-L axis and the femoral shaft axis. The inferior-superior axis of the femur was then defined by taking the cross product between the M-L and A-P axes, providing a mutually orthogonal set of anatomic axes for the femur (Fig 2). Positions of the femoral condyle centers measured in flexion and extension were expressed relative to the tibial origin. The angle of tibial rotation at each position was measured as the angle between the M-L axes of the femur and tibia projected onto the transverse plane of the tibia. The position of the femoral origin relative to the tibial origin was also recorded. Translations of the condyle centers, translation of the femoral origin, and amount of tibial rotation when moving from the extended position to the flexed position were computed by taking the differences between values in the flexed and extended positions. Measurements with a rotational phantom indicate an accuracy of 0.1° for tibiofemoral rotations. The
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION TABLE 1.
Internal Tibial Rotation in 9 Patients After ACL Reconstruction
Knee Position
ITR in ACL-Reconstructed Knees
ITR in Contralateral Knees
Extension 40° of flexion Extension to 40° of flexion
⫺4.3° ⫾ 4.7°* 5.3° ⫾ 3.7° 9.5° ⫾ 5.3°*
0.7° ⫾ 5.0°* 5.3° ⫾ 3.9° 6.0° ⫾ 6.0°*
NOTE. Data are given as mean ⫾ SD. Positions in full extension and approximately 40° of flexion are provided along with the amount of rotation that occurred when moving from the extended position to the flexed position. A positive value indicates internal tibial rotation, and a negative value indicates external tibial rotation. Abbreviation: ITR, internal tibial rotation. *Significant difference between ACL-reconstructed and contralateral knees (P ⬍ .05).
intraobserver and interobserver reproducibilities of the analysis techniques used in this study amount to 2 to 3 pixels (0.6 to 0.9 mm) for translations and approximately 1.5° for rotation.4 Cartilage-on-cartilage contact areas in the medial and lateral compartments were computed by connecting all spline points with a set of triangles and then summing the triangle areas. The contact centroid in each compartment was defined as the area centroid for the corresponding set of triangles. This method provided the overall center of contact for the compartment as a whole. Contact centroid coordinates were expressed with respect to the tibial origin, and centroid translations when moving from the extended position to the flexed position were computed as the difference between the locations in the 2 positions. For each measured parameter, a paired Student t test was performed to test for a significant difference (␣ ⫽ .05) between reconstructed knees and contralateral knees. Paired Student t tests were also used to detect significant changes in contact area and centroid location when moving from the extended position to the flexed position. RESULTS In the fully extended position (6.3° of hyperextension), the tibia was externally rotated by a mean of 4.3° in ACL-reconstructed knees, whereas the tibia in the contralateral limbs in the fully extended position (5.5° of hyperextension) was not significantly rotated relative to the femur (Table 1). In the flexed position, tibiae in ACL-reconstructed knees and contralateral knees were both internally rotated by a mean of 5.3°.
763
Because of the externally rotated position of the tibia in full extension, the tibia in ACL-reconstructed knees underwent a significantly larger amount of internal tibial rotation when moving from the extended position to the flexed position. Cartilage-on-cartilage contact area in ACL-reconstructed knees and contralateral knees decreased significantly in the medial compartment when moving from extension to flexion (Table 2). A decrease in contact area occurred in the lateral compartment in contralateral knees, but the contact area in the lateral compartment in ACL-reconstructed knees did not change significantly. When moving from extension to flexion, the contact centroid in the medial compartment underwent significant posterior translation in ACL-reconstructed and contralateral knees. The contact centroid in the lateral compartment underwent significant posterior and lateral translation in contralateral knees but not in ACL-reconstructed knees. No significant differences in contact area in either position or contact centroid translation were detected between ACL-reconstructed and contralateral knees. In the fully extended position, the A-P alignment of the whole femur in ACL-reconstructed knees was not significantly different than that in contralateral knees (Table 3). However, when the positions of the medial and lateral condyles were analyzed separately, the A-P
TABLE 2. Cartilage-on-Cartilage Contact Centroid Translations and Contact Areas in Patients After ACL Reconstruction ACL-Reconstructed Knees Translation of medial centroid (mm) A-P M-L Translation of lateral centroid (mm) A-P M-L Medial contact area (mm2) Extension Flexion Lateral contact area (mm2) Extension Flexion
5.5 ⫾ 3.0* 0.9 ⫾ 2.1
Contralateral Knees
6.2 ⫾ 2.1* 0.8 ⫾ 2.4
3.0 ⫾ 3.9 ⫺0.3 ⫾ 1.4
2.2 ⫾ 2.7* ⫺1.4 ⫾ 1.5*
308 ⫾ 77* 207 ⫾ 63
298 ⫾ 63* 189 ⫾ 40
174 ⫾ 52 159 ⫾ 69
195 ⫾ 45* 143 ⫾ 62
NOTE. Translations are expressed in the coordinate system of the tibia. Positive values indicate posterior translation and medial translation. *Significant change when moving from extension to flexion (P ⬍ .05).
764
R. D. CARPENTER ET AL. TABLE 3.
Side-to-Side Differences in Alignment of Femoral Condyles and Whole Femur in Fully Extended Position
Position in Full Extension A-P alignment of femur (mm) Internal tibial rotation (°) A-P condyle alignment (mm) Medial condyle Lateral condyle
ACL Reconstruction Patients (Reconstructed ⫺ Contralateral) (n ⫽ 9)
ACL-Deficient Patients (Injured ⫺ Contralateral) (n ⫽ 8)
1.0 ⫾ 2.3 ⫺3.6 ⫾ 4.2*†
2.6 ⫾ 1.7* 4.9 ⫾ 5.5*†
2.6 ⫾ 2.2* ⫺0.2 ⫾ 4.0†
1.4 ⫾ 2.9 3.8 ⫾ 2.0*†
NOTE. Data are given as mean ⫾ SD. A positive value indicates a more posterior alignment, a more medial alignment, or a more internally rotated position. Data for ACL-deficient knees are from Shefelbine et al.4 *Significant difference between affected and contralateral knees (P ⬍ .05). †Significant difference between ACL-deficient and ACL reconstruction patients (P ⬍ .05).
position of the medial femoral condyle in ACL-reconstructed knees in full extension was significantly (P ⬍ .05) posterior to that in contralateral knees whereas no difference was present on the lateral side. The relative positions of the medial and lateral condyles produced the externally rotated tibia position observed in the fully extended position. DISCUSSION The results of this study suggest that ACL reconstruction partially restores normal knee kinematics, but some important differences in kinematics between ACL-reconstructed and contralateral knees were observed. The kinematics of the lateral compartment were restored between extension and flexion, whereas those in the medial compartment were not, resulting in increasing internal tibial rotation between extension and flexion in the ACL-reconstructed knees. We compared our results with those in a previous cohort of ACL-deficient patients tested under similar loading conditions.4 Reconstructed knees were imaged at flexion angles of ⫺5.9° ⫾ 4.4° in the fully extended position and 39.7° ⫾ 12.1° in the flexed position. For comparison, the previously analyzed ACL-deficient patients were imaged at ⫺9.9° ⫾ 4.7° in the fully extended position and 42.7° ⫾ 8.3° in the flexed position. Because the mean angles in the 2 positions were not significantly different (P ⫽ .17 in extension and P ⫽ .60 in flexion) between the 2 study populations, comparisons between the 2 groups of patients were deemed reasonable. In the fully extended position the posterior alignment of the lateral condyle observed in ACL-deficient knees was restored to a normal position in ACLreconstructed knees, and normal A-P alignment of the
whole femur was also restored (Table 3). However, the center of the medial femoral condyle in ACLreconstructed knees was 2.6 mm posterior to that in contralateral knees in full extension (i.e., the medial side of the tibia was anteriorly subluxated). When ACL-reconstructed knees moved from the abnormal, externally rotated position in extension to a normally rotated position in flexion, ACL-reconstructed knees exhibited a total of 3.5° more internal tibial rotation than contralateral knees (Table 1). This result is similar to that obtained by Georgoulis et al.,20 who found that the tibia in ACL-reconstructed knees underwent approximately 4° more total rotation than normal knees during the stance phase of gait. Although the mechanical loads at the knee are very different between gait studies and our loading device, the 2 positions used in our study may capture a range of knee motion similar to that which occurs during the stance phase of gait. Our measurements of internal/external tibial rotation also correspond well with an in vivo study of running performed by Tashman et al.21 In that study ACL-reconstructed knees were slightly externally rotated (approximately 2° to 3°) at foot strike, which was also near the moment of maximum knee extension. Normal knees, on the other hand, were internally rotated by approximately 1° at foot strike. Then, as the knee flexed during the stance phase, the tibia rotated internally in both types of knees, reaching an internal rotation of 5° to 10°. Although the loading configuration and dynamic environment during running are very different from those of the simplified extension/ flexion configuration used in our study, our results (Table 1) show a range of tibial internal/external rotation and a difference between ACL-reconstructed
MRI OF KINEMATICS AFTER ACL RECONSTRUCTION and intact knees consistent with the results of Tashman et al. Our image-based method allowed analysis of the relative contributions of medial and lateral knee motion to knee rotational kinematics. Our results suggest that ACL reconstruction restores a normal amount of A-P motion on the lateral side of the knee but not on the medial side of the knee. This finding may have implications in understanding the effects of ACL injury and reconstruction on OA progression, which occurs predominantly in the medial compartment. Despite the difference in rotational kinematics, cartilage-on-cartilage contact areas and locations were not significantly different between ACL-reconstructed and contralateral knees, which is encouraging in terms of restoring a normal biomechanical environment to the ACL-injured knee. In the previous study of ACLdeficient knees performed in our laboratory, the locations of the medial and lateral contact centroids were posterior to those in contralateral knees.4 It appears that ACL reconstruction in our cohort restored contact centroids to their normal locations. The sizes of the contact regions were also not significantly different between ACL-reconstructed and contralateral knees, suggesting that the joint contact force is distributed over a similar area of cartilage in reconstructed knees. However, although the contact area and centroids may be close to normal, the abnormal initial tibial rotation and the resulting increase in internal tibial rotation when moving from extension to flexion may affect the distribution of loads in the ACL-reconstructed knee. For example, a study using finite-element models showed that 5° of increased internal tibial rotation after ACL injury leads to changes in cartilage stresses that may subsequently accelerate the rate of cartilage thinning.1 The loading configuration used in this study was intended to simulate partial weight bearing. Other studies have used dual-plane fluoroscopy22 and openMRI systems14 to analyze ACL-reconstructed kinematics during weight bearing. In patients in full knee extension during weight bearing, Papannagari et al.22 found that the tibia in the ACL-reconstructed knee was shifted by a mean of 2.9 mm anterior to that in contralateral knees. Although we did not observe this anterior shift for the whole tibia, we did see a similar shift (2.6 mm) on the medial side only. This finding again illustrates the value of analyzing the medial and lateral compartments separately. Also unlike Papannagari et al., we did not observe significant anterior translation of the tibia during flexion. It is likely that this difference is related to our use of a femur coor-
765
dinate system based on the centers of spheres fit to the femoral condyles instead of the transepicondylar axis. Logan et al.14 used images acquired with an open-MRI system to analyze medial and lateral femoral condyle motion during knee flexion using a single sagittal slice through each condyle. They found that the lateral side of the tibia in ACL-reconstructed knees was always shifted 5 mm in the anterior direction at each of 4 flexion positions from 0° to 90°. We did not find any significant differences in lateral femoral condyle motion between ACL-reconstructed and contralateral knees. This discrepancy could be because we used the positions of 3D sphere centers instead of circles on single sagittal slices to track condyle positions, or it may be a result of different surgical techniques. This study had some limitations that should be considered when interpreting the results. First, the loading configuration was not the same as that which occurs in full weight bearing or during functional activities such as walking. However, as stated previously, the 125-N compressive load used in this study was sufficient to produce a result (anterior subluxation of the tibia in extension) similar to that seen in full weight bearing with dual-plane fluoroscopy. In addition, although the amount of knee flexion observed in our study was similar to that which occurs during the stance phase of gait, space limitations prevented us from examining the knee at higher degrees of flexion, such as those that occur during squatting or step-up activities.23 It may be important for future studies to analyze kinematics at a higher angle of flexion than that used in our study. The use of 2 static knee positions to compute translations and rotations may also miss differences in knee motion that occur between these 2 positions. Finally, the ACL reconstructions were performed with a variety of graft and fixation choices, and it would be ideal to examine only a single graft type. However, the ACL reconstructions were performed by the same surgeon using the same femoral and tibial tunnel placements, minimizing surgeon and tunnel position variability. Future studies will include a larger sample size and investigate the effect of various ACL reconstructions such as doublebundle ACL reconstruction and medial portal drilling of femoral tunnels. CONCLUSIONS Reconstruction of the ACL restored normal motion on the lateral side of the knee but not on the medial side, resulting in increased internal tibial rotation when moving from full extension to 40° of flexion.
766
R. D. CARPENTER ET AL.
These results suggest that ACL reconstruction does not restore normal kinematics on the medial side of the knee, which may lead to early cartilage degeneration. Acknowledgment: The authors thank Sandra Shefelbine for the use of her data from ACL-deficient knees.
REFERENCES 1. Andriacchi TP, Briant PL, Bevill SL, Koo S. Rotational changes at the knee after ACL injury cause cartilage thinning. Clin Orthop Relat Res 2006;442:39-44. 2. Andriacchi TP, Mundermann A. The role of ambulatory mechanics in the initiation and progression of knee osteoarthritis. Curr Opin Rheumatol 2006;18:514-518. 3. Defrate LE, Papannagari R, Gill TJ, Moses JM, Pathare NP, Li G. The 6 degrees of freedom kinematics of the knee after anterior cruciate ligament deficiency: An in vivo imaging analysis. Am J Sports Med 2006;34:1240-1246. 4. Shefelbine SJ, Ma CB, Lee KY, et al. MRI analysis of in vivo meniscal and tibiofemoral kinematics in ACL-deficient and normal knees. J Orthop Res 2006;24:1208-1217. 5. Li G, Papannagari R, DeFrate LE, Yoo JD, Park SE, Gill TJ. The effects of ACL deficiency on mediolateral translation and varus-valgus rotation. Acta Orthop 2007;78:355-360. 6. Andriacchi TP, Mundermann A, Smith RL, Alexander EJ, Dyrby CO, Koo S. A framework for the in vivo pathomechanics of osteoarthritis at the knee. Ann Biomed Eng 2004;32:447457. 7. Church S, Keating JF. Reconstruction of the anterior cruciate ligament: Timing of surgery and the incidence of meniscal tears and degenerative change. J Bone Joint Surg Br 2005;87: 1639-1642. 8. Ohly NE, Keating JF. Arthrosis following ACL reconstruction. In: Prodromos C, Brown C, Fu FH, Georgoulis AD, eds. The anterior cruciate ligament: Reconstruction and basic science. Philadelphia: Saunders, 2008;69-79. 9. Lohmander LS, Ostenberg A, Englund M, Roos H. High prevalence of knee osteoarthritis, pain, and functional limitations in female soccer players twelve years after anterior cruciate ligament injury. Arthritis Rheum 2004;50:3145-3152. 10. Maletius W, Messner K. Eighteen- to twenty-four-year follow-up after complete rupture of the anterior cruciate ligament. Am J Sports Med 1999;27:711-717.
11. Asano H, Muneta T, Ikeda H, Yagishita K, Kurihara Y, Sekiya I. Arthroscopic evaluation of the articular cartilage after anterior cruciate ligament reconstruction: A short-term prospective study of 105 patients. Arthroscopy 2004;20:474-481. 12. Johal P, Williams A, Wragg P, Hunt D, Gedroyc W. Tibiofemoral movement in the living knee. A study of weight bearing and non-weight bearing knee kinematics using ‘interventional’ MRI. J Biomech 2005;38:269-276. 13. Logan M, Dunstan E, Robinson J, Williams A, Gedroyc W, Freeman M. Tibiofemoral kinematics of the anterior cruciate ligament (ACL)-deficient weightbearing, living knee employing vertical access open “interventional” multiple resonance imaging. Am J Sports Med 2004;32:720-726. 14. Logan MC, Williams A, Lavelle J, Gedroyc W, Freeman M. Tibiofemoral kinematics following successful anterior cruciate ligament reconstruction using dynamic multiple resonance imaging. Am J Sports Med 2004;32:984-992. 15. Scarvell JM, Smith PN, Refshauge KM, Galloway HR, Woods KR. Evaluation of a method to map tibiofemoral contact points in the normal knee using MRI. J Orthop Res 2004;22:788-793. 16. Scarvell JM, Smith PN, Refshauge KM, Galloway HR, Woods KR. Association between abnormal kinematics and degenerative change in knees of people with chronic anterior cruciate ligament deficiency: A magnetic resonance imaging study. Aust J Physiother 2005;51:233-240. 17. Scarvell JM, Smith PN, Refshauge KM, Galloway HR. Magnetic resonance imaging analysis of kinematics in osteoarthritic knees. J Arthroplasty 2007;22:383-393. 18. Lee KY, Slavinsky JP, Ries MD, Blumenkrantz G, Majumdar S. Magnetic resonance imaging of in vivo kinematics after total knee arthroplasty. J Magn Reson Imaging 2005;21:172178. 19. Besl PJ, McKay ND. A method for registration of 3-D shapes. IEEE Trans Pattern Anal Mach Intell 1992;14:239-256. 20. Georgoulis AD, Ristanis S, Chouliaras V, Moraiti C, Stergiou N. Tibial rotation is not restored after ACL reconstruction with a hamstring graft. Clin Orthop Relat Res 2007;454:89-94. 21. Tashman S, Collon D, Anderson K, Kolowich P, Anderst W. Abnormal rotational knee motion during running after anterior cruciate ligament reconstruction. Am J Sports Med 2004;32: 975-983. 22. Papannagari R, Gill TJ, Defrate LE, Moses JM, Petruska AJ, Li G. In vivo kinematics of the knee after anterior cruciate ligament reconstruction: A clinical and functional evaluation. Am J Sports Med 2006;34:2006-2012. 23. Andriacchi TP, Dyrby CO. Interactions between kinematics and loading during walking for the normal and ACL deficient knee. J Biomech 2005;38:293-298.