Failure of Reconstruction of the Anterior Cruciate Ligament Due to Impingement by the Intercondylar Roof*† BY MAJOR STEPHEN M. HOWELL‡ MEDICAL CORPS, UNITED STATES AIR FORCE RESERVE, AND CAPTAIN MICHAEL A. TAYLOR§, MEDICAL SERVICE CORPS, UNITED STATES AIR FORCE

The relationship between impingement of the roof of the intercondylar notch on a reconstructed anterior cruciate ligament, and the subsequent stability and range of extension of the joint, was analyzed in fortyseven knees. The extent of the impingement was determined by analysis of the relationship of the tibial tunnel to the intersection of the line of slope of the intercondylar roof with the plane of the subchondral bone of the articular surface of the tibial plateau. These lines were drawn on a lateral roentgenogram that was made with the knee in maximum extension, two years after the operation. In all four knees in which the entire articular opening of the tibial tunnel was anterior to the slope of the intercondylar roof, there was severe impingement on the graft, and all four grafts failed. In the fourteen knees in which a portion of the articular opening of the tibial tunnel was anterior to the slope of the intercondylar roof, there was moderate impingement on the graft, and four grafts failed (an unacceptable rate of failure). There was no impingement in the knees in which the entire articular opening of the tibial tunnel was posterior to the slope of the intercondylar roof, and these knees were associated with the lowest rate of failure of the grafts (three of twenty-nine). Knees that had an impinged graft and regained a complete range of extension became unstable. Results from a biomechanical study suggested that the placement of the tibial tunnel may be important for the clinical success of reconstruction of the anterior cruciate ligament. A graft that is placed in a tibial tunnel that is anterior to the slope of the intercondylar roof is subject to *One or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund or foundation, educational institution, or other non-profit organization with which one or more of the authors are associated. No funds were received in support of this study. †The views expressed herein are those of the authors and do not reflect the official policy or position of the United States Department of Defense or the United States Government. ‡7601 Timberlake, Suite 103, Sacramento, California 95823. Please address requests for reprints to Dr. Howell. §Department of Orthopedics (SGHT), David Grant Medical Center, Travis Air Force Base, California 94535-5300. higher tensile and compressive loads than is a graft in a tibial tunnel that is posterior to the slope of the intercondylar roof. Increased loads can occur as the knee nears full extension because of impingement of the intercondylar roof on an anteriorly placed graft. Thus, the clinical result may be related to the placement of the tibial tunnel and the severity of the impingement. Impingement by the roof substantially changes the appearance of an anterior cruciate-ligament graft on a magnetic resonance image, There is a predictable increase in the signal intensity of grafts that have been placed anteriorly, apparently because of the impact of the roof on the anterior surface of the graft during extension of the knee. Grafts that have been placed in tibial tunnels that are aligned posterior and parallel to the slope of the intercondylar roof retain a low signal intensity. The purpose of this study was to determine the relationship between impingement on the graft by the roof and the stability and range of extension of the knee, and to assess whether there was any association be-tween the slope of the intercondylar roof, the placement of the tibial tunnel, or the range of extension or the stability of the knee two years after the reconstruction. Methods and Materials Selection of Patients Group I comprised twenty-six patients who had been operated on consecutively between December 1986 and December 1987. The senior one of us (S. M. H.) was either the primary surgeon or the first assistant at these operations. An extra-articular reconstruction was in-cluded in the procedure. Three patients were excluded from the study because intraoperative testing for isometry had not been done for them, and four patients were lost to follow-up. Thus, nineteen patients from Group I were included in this study. Group II comprised thirty-eight patients who had been operated on consecutively between March 1989 and March 1990. The senior one of us was the principal surgeon at these operations. For the patients in Group II, the orientation of the tibial tunnel was directed on the basis of an intraoperative roentgenogram, and the intercondylar notch was enlarged to accommodate the volume of the graft when 1044 THE JOURNAL OF BONE AND JOINT SURGERY FAILURE OF RECONSTRUCTION OF THE ANTERIOR CRUCIATE LIGAMENT DUE TO IMPINGEMENT 1045 VOL. 75-A, NO. 7, JULY 1993 the knee was in maximum extension. An extra-articular reconstruction was not done for the patients in Group II. Ten Group-II patients were excluded from the study: eight patients could not return for follow-up, one patient had a chronic rupture of the anterior cruciate ligament in the contralateral knee, and the graft in one patient ruptured when the patient fell from a height of three meters. Thus, twentyeight patients from Group II were included in this study. Lateral roentgenograms of the knee were made at least two years after reconstruction of the anterior cruciate ligament. The roentgenograms were made with the patient supine and the heel on the side of the reconstructed knee elevated on a foam bolster, so that the popliteal fossa was suspended ten centimeters above the table. The patient was instructed to relax the lower limb, to allow gravity to pull the knee into maximum extension. The x-ray beam was directed medially, parallel to the joint line. Roentgenograms were made repeatedly until the lateral projections of the medial and femoral condyles were superimposed (Fig. l-A). Measurements were made from roentgenograms on which the offset of the overlap of the medial and lateral femoral condyles was six millimeters or less (Figs. l-A, 2-A, 3-A, 4-A, and 4-B) Impingement by the roof was assessed on the lateral roentgenogram by study of the relationship of the tibial tunnel to the point of intersection of the line of the slope of the intercondylar roof with the plane of the articular surface of the tibial plateau (Figs. l-A, 2-A, 3-A, 4-A, and 4-B). The plane of the tibial plateau was defined by a line between the most superior points of the anterior and posterior margins of the proximal end of the tibia. Impingement was considered to be severe when the posterior border of the superior end of the tibial tunnel (or its extension to the plane of the tibial plateau) was at, or anterior to, the intersection of the line of the slope of the intercondylar roof with the plane of the tibial plateau (Fig. 1-A). Impingement was considered to be moderate when a portion of the tibial tunnel was anterior to this intersection (Figs. 2-A, 4-A, and 4-B). No impingement was considered to be present when the anterior border of the tibial tunnel was at, or posterior to, this intersection (Fig. 3-A). The amount of impingement by the roof was calculated by measurement, on the lateral roentgenogram, of the distance on the line of the tibial plateau from the point where the line of the anterior edge of the tibial tunnel intersected the plateau, to the point where the line of the slope of the intercondylar roof intersected the plateau. This distance was then divided by the width of the tibial tunnel, and the result was expressed as a percentage. A positive ratio was used to indicate impingement and a negative one, no impingement. For example, the percentage of roof impingement was calculated to be 167 per cent in Figure l-A (severe impingement), 25 per cent in Figure 2-A (moderate im-pingement),0 per cent in Figure 3-A (no impingeFIG. 1-A FIG l-B Figs. 1-A and l-B: A knee in which there was severe impingement on the anterior cruciate-ligament graft. The graft had been placed in a tibial tunnel, the entire width of which was anterior to the point of intersection of the slope of the intercondylar roof with the plane of the articular surface of the tibial plateau. Fig. 1-A: Lateral roentgenogram. Roof impingement was 167 per cent, and the angle of the roof was 36 degrees. f = longitudinal axis of the femur, r = slope of intercondylar roof, and o = plane of tibial plateau. Fig. 1-B: Sagittal magnetic-resonance image (repetition time, 1200 milliseconds; echo time, forty milliseconds). There was a high signal intensity in the graft that was diagnostic of roof impingement. The graft had been angulated and elongated by impingement from the intercondylar roof, but it appeared to be in continuity and was not completely ruptured. S.M. HOWELL AND M.A. TAYLOR 1046 THE JOURNAL OF BONE AND JOINT SURGERY ment), 20 per cent in Figure 4-A (moderate impingement), and 54 per cent in Figure 4-B (moderate impingement). The location of the central axis of the tibial tunnel was calculated by extension of the line of the central axis of the tibial tunnel to its intersection with the line of the tibial plateau: the distance from this intersection to the anterior end of the line of the tibial plateau was then measured. This distance was divided by the length of the line of the tibial plateau, and the result was expressed as a percentage The slope of the intercondylar roof was measured as the angle subtended by the line of the slope of the roof with the line of the long axis of the femur (Fig. 1-A). Four patients in Group I had severe impingement: thirteen, moderate impingement; and two, no impingement. No patient in Group II had severe impingement, one had moderate impingement, and twenty-seven had no impingement. Effect of Flexion Contracture on Classification of Impingement on the Graft The classification of severe, moderate, or no impingement on the graft was based on the percentage of roof impingement that was determined from the lateral roentgenogram of the maximally extended knee. Seventeen of the reconstructed knees had a flexion contracture, and fifteen of the seventeen had an extension deficit of 5 degrees or less. The relationship of the tibial tunnel to the point of intersection of the line of the slope of the intercondylar roof with the plane of the tibial plateau can be expected to be more posterior in a knee that is flexed than in a knee that is fully extended. It is possible that the initial calculation and classification of the severity of the impingement on the graft may have been underestimated in the knees in which there was a flexion contracture. The following analysis was performed to estimate the increase in the severity of impingement that could be expected if a knee in which there was a flexion contracture regained full extension. A lateral roentgenogram, made with the knee in maximum extension, was obtained for four volunteers who had normal knees: a man who was twenty-two years old, and three women who were twenty-two, thirty-six, and fiftytwo years old (Figs. 4-A and 4-B). A foam bolster was then placed posterior to the thigh and was adjusted proximally or distally to create the position of a flexion contracture. Additional roentgenograms were made for each subject, with the knee in 5 and 10 degrees of flexion. The locations of the tibial tunnel that would result in severe, moderate, and no impingement on the graft were represented on each roentgenogram by three nine-millimeter-wide tibial tunnels, drawn centrally at 22, 32, and 42 per cent of the disF IG. 2-A F IG. 2-B Figs. 2-A and 2-B: A knee in which there was moderate impingement on the graft. Fig. 2-A: Lateral roentgenogram. A portion of the tibial tunnel was anterior to the intercondylar roof. There was 25 per cent roof impingement, and the angle of the roof was 39 degrees. Fig. 2.B: Sagittal magnetic-resonance image (repetition time, 1200 milliseconds: echo time, forty milliseconds). A ruptured anterior bundle can be seen. There is no longer impingement on the ruptured bundle because it had been displaced into the anterior chamber of the knee, and it had a low signal intensity, characteristic of a graft on which there is no impingement. The remaining portion of the posterior bundle had a regionalized increase in the magnetic resonance signal, characteristic of impingement on a graft (arrow). (A delayed enlargement of the roof was performed later, and the posterior portion of the graft that appeared to be absent on the magnetic resonance image was found to be intact on probing. The knee remained stable two years after the initial operation.) tance from the anterior to the posterior edge of the tibial plateau. The percentage of roof impingement was calculated and used to determine the increase in the percentage that may result when the knee is moved in 5-degree increments from 10 degrees of flexion to full hyperextension.