You are in: eMedicine Specialties > Radiology > MUSCULOSKELETAL Knee, Anterior Cruciate Ligament Injuries (MRI)Article Last Updated: Jun 14, 2006AUTHOR AND EDITOR INFORMATIONSection 1 of 11
  Author: Anton M Allen, MD, Assistant Residency Program Director, Associate Professor, Department of Radiology, University of Tennessee Medical Center at Knoxville Anton M Allen is a member of the following medical societies: American College of Radiology, American Roentgen Ray Society, Radiological Society of North America, and Society of Skeletal Radiology Coauthor(s): Alan W Horn, MD, Consulting Staff, Department of Radiology, Bryan Radiology Associates; Timothy N Ozburn, MD, Staff Physician, Department of Radiology, University of Tennessee Medical Center at Knoxville Editors: David S Levey, MD, PhD, Orthopedic/Spine MRI TeleRadiologist, Radsource, LLC; Bernard D Coombs, MB, ChB, PhD, Consulting Staff, Department of Specialist Rehabilitation Services, Hutt Valley District Health Board, New Zealand; Javier Beltran, MD, Chair, Department of Radiology, Maimonides Medical Center; Robert M Krasny, MD, Consulting Staff, Department of Radiology, The Angeles Clinic and Research Institute; Felix S Chew, MD, MBA, EdM, Professor, Department of Radiology, Vice Chairman for Radiology Informatics, Section Head of Musculoskeletal Radiology, University of Washington Author and Editor Disclosure Synonyms and related keywords: anterior cruciate ligament tear, anterior cruciate ligament sprain, ACL tear, ACL sprain, ACL injury, anterior cruciate ligament MR, ACL MR, anterior cruciate ligament MRI, ACL MRI, anterior cruciate ligament magnetic resonance imaging, ACL magnetic resonance imaging, imaging of the ACL, imaging of the anterior cruciate ligament, MRI of the anterior cruciate ligament, MR of the anterior cruciate ligament, magnetic resonance imaging of the anterior cruciate ligament, imaging of ACL injuries, imaging of anterior cruciate ligament injuries INTRODUCTIONSection 2 of 11
  The anterior cruciate ligament (ACL) is the most commonly injured of the major knee ligaments. Injuries occur frequently in both athletes and nonathletes. The ACL is a vital ligamentous stabilizer of the knee that resists anterior translation and secondarily resists varus and valgus forces (Swenson, 1995). The ACL also functions as a mechanoreceptor that relays information about knee tension to the central nervous system. Patients with ACL injury have variable knee instability that may limit even ordinary daily activities. They report particular difficulty with cutting and pivoting. The torn ACL undergoes limited healing. Long-term morbidity is common with sequelae including injury to the articular cartilage, secondary meniscal tears, and osteoarthritis. Diagnosis ACL injury is usually diagnosed on the basis of the patient's history and physical findings or MRIs. Arthroscopy and arthrotomy are the criterion standards for diagnosis, but they are invasive and costly. The skilled clinician can diagnose as many as 90% of ACL tears based on history and physical examination findings (Johnson, 1993; Lee, 1998). Patients typically report an audible pop and "giving way" at the time of injury. A knee effusion usually develops over the next 24 hours. A tear is confirmed by physical examination, primarily by performing the Lachman test (Swenson, 1995). The anterior drawer and pivot shift tests are often helpful, and arthrometric examination may be contributory. Physical diagnosis may be difficult in large patients, in patients with strong secondary muscular restraints, and in patients with an acute injury and soft tissue swelling and guarding. Partial ACL tears are especially difficult to diagnose on physical examination (Noyes, 1989). MRI may provide pivotal diagnostic information about the ACL in all of these settings (Otani, 2001; Munk, 1998). Role of MRI As noted above, MRI may alter the treatment by allowing confident diagnosis or exclusion of an ACL tear in patients with equivocal physical examination findings; however, the greatest contribution of MRI in ACL-injured patients is in the evaluation of coexisting internal derangements (Munk, 1998). Diagnosis or exclusion of these comorbidities often directs treatment. Four specific examples are discussed below. The first example is posterolateral-corner injury. The posterolateral support structures are many but include the fibular collateral, biceps femoris ligament, popliteus ligament, and the posterolateral capsule (Veltri, 1994). Injury to these structures can easily be missed upon clinical examination and arthroscopy (Hughston, 1985; Veltri, 1994), but they are often detected on MRIs (Miller, 1997). Knees with tears to both the ACL and lateral collateral ligament (LCL) are usually markedly unstable. Furthermore, unrepaired posterolateral-corner injuries predispose ACL grafts to early failure (Hughston, 1985). As such, MRI diagnosis of coexisting ACL/LCL tears indicates the need for surgery to both structures. In addition, surgery is often performed earlier than it would be if an LCL tear were not present. This is because most orthopedic surgeons favor delaying ACL surgery until range of motion is optimized, while posterolateral-corner repairs are optimally performed early ( <3 wk) (White, 2004). Second are posterior cruciate ligament (PCL) tears. If both PCL and ACL tears are present, instability is usually profound, and reconstructive surgery of both ligaments is ordinarily necessary. PCL tears are readily diagnosed by using MRI, but they can be difficult to detect on physical examination, even those performed by experienced examiners. Third are meniscal tears. Meniscal repairs have a higher rate of failure in ACL-deficient knees than in ACL-reconstructed knees (Cannon, 1992). Therefore, ACL surgery is more likely to be recommended in patients who are undergoing meniscal repair. In addition, MRI diagnosis of a displaced bucket-handle tear indicates a need for early arthroscopic surgery, especially if knee locking is present (Veltri, 1994). Be especially careful to inspect the menisci in patients with ACL injury because sensitivity for diagnosis of meniscal tear significantly decreases in this setting (Won-Hee-Jee, 2004). Fourth are extensor-mechanism abnormalities. MRI-enabled diagnosis of clinically significant patellofemoral chondromalacia, patellofemoral osteochondral fractures, or other extensor derangements, may mitigate against patellar-autograft ACL reconstruction. Instead, patellar allograft or alternative autograft reconstructions may be elected (personal communication). Although diagnosis of specific internal derangements alters treatments in patients with ACL tears, Vincken et al reported that the evaluation of the joint as a whole (composite diagnosis) is relevant in selecting patients for therapeutic arthroscopy. The sensitivity of MRI composite diagnosis is 87-94%, and its specificity is 88-93% (Vincken, 2002). Owing to its high composite sensitivity, MRI can significantly decrease the number of unnecessary endoscopy procedures performed (Munk, 1998). Treatment Treatment of ACL tears ranges from conservative therapies to surgical ACL reconstruction. The patient's activity level (and expectations for activity in the future) is the most important factor guiding the choice of treatment (Swenson, 1995). Associated meniscal and ligamentous injuries, the degree of laxity, and the patient's age and willingness to pursue vigorous postoperative physical therapy are other major determinants. ACL graft reconstruction stabilizes the ACL-deficient knee, increasing activity levels and preventing reinjury from repeated subluxation. ACL reconstruction, however, has not yet been proven to prevent long-term osteoarthritic deterioration (Swenson, 1995; Dye, 1998). It is generally believed that late ACL reconstruction decreases postprocedural stiffness and improves outcomes. Surgery is delayed until swelling has subsided and range of motion is restored (Swenson, 1995). For excellent patient education resources, visit eMedicine's Foot, Ankle, Knee, and Hip Center and Sprains and Strains Center. Also, see eMedicine's patient education articles Knee Injury, Knee Pain, and Magnetic Resonance Imaging (MRI). ANATOMYSection 3 of 11
The ACL is a dense fibrous band composed of collagen fibrils. It is about 3.5 cm long and 1 cm in transverse diameter (Resnick, 1995). It originates from the posteromedial aspect of the lateral femoral condyle and courses through the lateral intercondylar notch in an anterior, inferior, and medial direction. It inserts on the tibia approximately 23 mm posterior to the anterior edge of the tibia, just anterior and lateral to the medial intercondylar eminence (tibial spine) (Resnick, 1995; Stoller, 1997). The ACL is not as strong as the PCL and it is less strong at its femoral origin than at its tibial insertion (Resnick, 1995). The ACL is organized into an indeterminate number of linear fascicles that are often partially visible on MRI. The fascicles diverge (fan) distally into a larger foot-like insertion on the tibia that facilitates tucking of the ACL under the anterior femoral intercondylar roof (Resnick, 1995). ACL fascicles are organized into functional anteromedial and posterolateral bundles or bands (Girgis, 1975) that are named for their location relative to each other at tibial insertion (Resnick, 1995). These bundles are twisted about each other and cannot be differentiated on MRIs. The strong anteromedial bundle tightens with flexion of the knee and probably resists anterior translation of the tibia in flexion. The posterolateral bundle tightens with knee extension and probably resists hyperextension (Resnick, 1995). The physiologic property in which part of the spiraled ACL is taut throughout the normal range of motion of the knee is termed isometry. Graft isometry is one goal of reconstructive surgery (though it is probably uncommonly achieved). The ACL is an extrasynovial and intracapsular ligament. Bands of mesenterylike synovium, arising from the posterior intercondylar region of the tibia, surround the cruciate ligaments (Resnick, 1995). This feature accounts for fluid often seen anterior to the normal ACL (and posterior to the PCL) on MRI. The extrasynovial location also helps to explain why hemarthrosis may be delayed with an acute ACL tear. The primary blood supply to the ACL derives from arteries to the surrounding synovial membrane. These in turn derive from branches of the middle geniculate artery piercing the posterior capsule (Resnick, 1995). The central core of the ACL is relatively avascular. This may partly account for the generally ineffective healing of ACL tears. Tibial nerve terminal branches innervate the ACL (Resnick, 1995). MECHANISM OF INJURYSection 4 of 11
Mechanisms of ACL injury are numerous; they are discussed here only superficially. ACL tears occur with or without contact, and with the knee in any position from flexed to fully extended. The most common contact mechanism of injury is the valgus-abduction clip injury (Stoller, 1997). These injuries are common in football players and occur with a lateral blow to the partially flexed knee. Coexisting medial and lateral meniscal tears are common, as are medial collateral ligament (MCL) injuries. Hyperextension or varus-hyperextension from an anterior blow (eg, injury from a motor vehicle accident or contact sports) is the second most common contact mechanism of ACL injury. The PCL and/or posterolateral-corner structures are also frequently injured. With more severe hyperextension, the knee may dislocate; the popliteal neurovascular bundle or peroneal nerve may be injured in this setting. Noncontact mechanisms account for 70-80% of ACL tears (DeMorat, 2004; Stoller, 1997). The pivot-shift mechanism is most commonly implicated: the slightly flexed knee incurs a valgus load, with internal rotation of the tibia or external rotation of the femur. This twisting injury often occurs with rapid simultaneous deceleration and directional movements in skiers, football, basketball, or soccer players. Marked quadriceps loading at the time of injury has been implicated (DeMorat, 2004). Associated meniscal tears, collateral ligament injuries, and lateral patellar subluxation are common. Noncontact hyperextension, such as that occurring in a gymnast or cheerleader who misses a landing, is another mechanism of injury that may result in ACL injury (Stoller, 1997). The incidence of ACL tears in females is increased 2-10 times over that of males for each hour of participation in activities at risk (Gwinn, 2000; Arendt, 1999). Explanations for this increased susceptibility are under debate (Lovering, 2005). Lax joints may also predispose patients to ACL injury (Ramesh, 2005). MRI TECHNIQUE FOR ACL EVALUATIONSection 5 of 11
Relevant history and physical examination should be included in the clinical information submitted with the patient. Information especially helpful to the radiologist includes history regarding previous knee surgeries and dates of injuries. The authors have found it beneficial for technologists to place MRI markers at sites of pain and surgical scars. The reader is referred elsewhere for the basic principles of generating high-quality MRIs. Imaging protocols Knee MRI protocols must be designed to yield diagnostic images of not only the ACL but also of the menisci, bones, articular cartilage, and other ligamentous structures of the knee. Furthermore, the requirements for good meniscus and cartilage imaging are more exacting than the requirements for diagnostic ACL imaging. Therefore, for the most part, a protocol that images the menisci and cartilage optimally also demonstrates the ACL adequately. This explains why most centers image patients in full knee extension, though the ACL is optimally evaluated with the knee in about 30° of flexion. Imaging in flexion complicates evaluation of the menisci and other knee structures (Lee, 2004). The protocol requirements of ACL imaging are simply sequences in all 3 planes (sagittal, coronal, axial) (Fizgerald, 1993; Remer, 1992) that include both T1-weighted (or proton density–weighted) and T2-weighted sequences in the sagittal plane. Although the sagittal imaging plane is often most helpful in evaluating the ACL, any of the 3 imaging planes may prove pivotal in a given case. Axial images provide a unique cross-sectional view free of partial volume artifact with the intercondylar roof (Fitzgerald, 1993; Roychowdhury, 1996; Roychowdhury, 1997; Lerman, 1995) and are invaluable in evaluation of the proximal ACL. Coronal imaging is useful for evaluation of proximal and midportion tears (Remer, 1992). The radiologist should routinely inspect the ACL in all planes and become familiar with the range of normal and abnormal appearances in each plane. Other technical considerations The ACL is usually seen to greatest advantage on T2-weighted images as opposed to T1-weighted or gradient-echo images obtained with short echo times (Lee, 1998; Mink, 1988). This is partly due to confounding increased signal intensity seen in ligaments and tendons with short–echo time sequences owing to magic-angle effect and other factors. Fast spin-echo (termed turbo spin-echo by some vendors) T2-weighted sequences with fat saturation are performed faster than conventional T2-weighted sequences, and they have largely replaced them in centers that can perform these sequences. Several methods for the prescription of sagittal images of the ACL by technologists have been reported. Early recommendations were to allow patients to naturally externally rotate their legs and then to prescribe longitudinal images perpendicular to the table. However, this method leads to inconsistent results from patient to patient, and sometimes leads to overly oblique sagittal images that distort the meniscal anatomy. Best results are obtained if the technologist draws a bicondylar line that intersects the posterior margins of the medial and lateral femoral condyles on an axial scout image. A sagittal oblique imaging plane is then prescribed at an angle 10-15° to a perpendicular to this line. Whatever method is used, excessive off-sagittal obliquity should be avoided. Additional sequences Additional problem-solving imaging sequences of the ACL are rarely necessary. Katahira et al (2001) reported increased diagnostic accuracy prescribing oblique coronal images parallel to the long axis of the ACL off of an oblique sagittal image obtained as described above (double-oblique sequence). Several other investigators have reported similar findings (Staeubli, 1999; Hong, 2003). Investigators have also reported improved ACL visualization in the knee in mild (17-30°) flexion (Niitsu, 1998; Lee, 2004) because of decreased partial voluming of the proximal ACL with the intercondylar roof. Kinematic imaging for diagnosis of ACL tears has been proposed (Niitsu, 1991). None of the ancillary methods of ACL imaging described above has found wide application. The present authors suggest they only be used for occasional problem solving in cases of equivocal ACL findings. CT of the ACL MRI cannot be safely performed in some clinical settings such as in patients with pacemakers. Carefully performed CT arthrography is an alternative in this setting. Vande Berg et al (2002) reported 90% sensitivity and 96% specificity for diagnosis of ACL tear with CT arthrography. Submillimeter-resolution CT imaging was performed by using dual-detector spiral CT after an injection of contrast agent into the knee. Three-dimensional (3D) virtual CT has also been described (Irie, 2002). NORMAL ACL MRI APPEARANCESection 6 of 11
Normal MRI appearances On sagittal images, the normal ACL appears as a solid band or as a striated band diverging slightly distally. As many as 4 striations may be present (Remer, 1992). The normal ACL is usually ruler-straight, though mild sagging convex inferiorly can be present, especially with mild knee flexion. The ACL shows low-to-intermediate signal intensity, higher than that of the PCL. The distal ACL demonstrates relatively increased signal intensity, presumably due, in part, to distal divergence of fascicles. Data from one study confirmed that increased internal signal intensity is the result of macroscopic (rather than histologic) features of the ACL and that in elderly patients, internal degeneration accounts for some of the observed increased signal intensity (Hodler, 1992). On coronal images, normal ACL is usually well seen, though fascicles often appear attenuated and single or few in number. The lateral position of the ACL in the intercondylar notch of the femur is apparent in the coronal plane; the PCL is seen medially. In the axial plane, the proximal ACL is well seen and appears as an elliptical low signal intensity band adjacent to the lateral wall of the upper intercondylar notch. It gradually moves away from the wall and splits into a horseshoe (fan-shaped) array of fascicles as it approaches its tibial insertion (Roychowdhury, 1996). The distal ACL is difficult to critically evaluate on axial images. Pitfalls in interpreting normal findings The ACL is poorly visualized in 5-19% of healthy knees in the sagittal plane (Remer, 1992). T1-weighted or gradient-echo sagittal images are especially likely to demonstrate the ACL poorly. However, the absence of hemorrhage or edema in the expected location of the ACL, a normal appearance of the ACL in other planes, and the absence of secondary signs of ACL injury is usually sufficient to confirm that the ACL is normal (Remer, 1992). Smith et al (1996) noted that the ACL is almost always intact when it is not visualized on either T1- or T2-weighted sagittal MRIs, but is normal in appearance on images obtained with the other sequences. Partial-volume superimposition of the inner aspect of the lateral femoral condyle on the proximal ACL may produce a pseudomass that mimics an acute ACL tear on sagittal images. If section thicknesses of 4 mm or less are routinely used and if other imaging planes are correlated, this is not a diagnostic problem (Remer, 1992). The proximal origin of the ACL is often less well seen than the remainder of the ACL on sagittal images, partly because of its proximity to the adjacent intercondylar roof; however, this part of the ACL is usually well demonstrated on axial images. A repeat sagittal sequence with approximately 30° flexion of the knee could also be performed (Lee, 2004). MRI FINDINGS IN ACUTE INJURYSection 7 of 11
Most ACL tears (about 70%) occur in the middle aspect of the ligament (Resnick, 1995); 7-20% occur proximally near its origin. Only 3-10% occur distally at the tibial attachment (Remer, 1992; Resnick, 1995). Studies report 92-100% sensitivity and 82-100% specificity of MRI for the diagnosis of ACL tears (Robertson, 1994; Mink, 1988; Fitzgerald, 1993; Brandser, 1996; Lee, 1988; Pope, 1993; Tung, 1993; Bari, 2003; Winters, 2005). Accuracy is not affected significantly by field strength (Vellet, 1995). Sensitivity is significantly decreased if other major ligamentous injuries are present in the knee (Rubin, 1998). Data for children are less than that for adults. Decreased accuracy of MRI has been reported in preadolescents (McDermott, 1998), but a study of patients aged 5-16 years demonstrated a sensitivity of 95% and a specificity of 88% (Lee, 1999). Primary signs of ACL tear Primary signs of acute ACL tear (ie, MRI abnormalities of the ACL proper) allow for high accuracy in the diagnosis of ACL injury, even in the absence of secondary signs (Brandser, 1996; Lee, 1998; Mink, 1988; Tung, 1993; Falchook, 1996). The primary signs of an ACL tear include nonvisualization, disruption of the substance of the ACL by abnormal increased signal intensity, abrupt angulation or a wavy appearance, and an abnormal ACL axis. The axis of the ACL is abnormal if it is clearly more horizontal than a line projected along the intercondylar roof (Blumensaat line) on sagittal images. The ACL axis can be quantitated (although the authors have not found this to be necessary); a less than 45° angle of the long axis of the ACL relative to a line parallel to the tibial plateau (the ACL angle) is reportedly sensitive and specific for ACL tear (Mellado, 2004). A common presentation of an acute ACL tear is nonvisualization or near-nonvisualization, with replacement by a cloud of focal edema and hemorrhage. An acute tear manifesting as enlargement of the ACL and increased internal signal intensity but with visible intact fascicles has been termed an interstitial tear. This finding should be differentiated from mucoid degeneration of the ACL. Primary signs of tear involving the proximal ACL should be sought on axial images. The linear hypointense band representing the proximal ACL may be attenuated, completely or partially replaced by hemorrhage, or displaced away from the lateral wall of the intercondylar notch (Roychowdhury, 1997). Secondary signs of ACL tear MRI findings of an ACL tear apart from abnormalities of the ACL proper are termed secondary signs. The sensitivity of these signs is limited (Brandser, 1996); therefore, the absence of secondary signs in no way excludes ACL disruption. Certain signs, however (discussed below), have greater than 80% specificity. As a consequence, they may allow for a confident diagnosis of tear when primary signs are equivocal (Vahey, 1991 and 1993; Cobby, 1992; Kaplan, 1992; Murphy, 1992; Tung, 1993; Chan, 1994; Gentili, 1994; McCauley, 1994; Robertson, 1994; Branser, 1996; Kaplan, 1999). Secondary signs with high specificity for ACL injury include pivot-shift bone bruises/osteochondral fractures, anterior translocation of the tibia, and Segond fractures. Pivot-shift bone bruises and fractures Characteristic posterolateral tibial plateau and lateral femoral condyle subchondral bone bruises occur with the pivot-shift mechanism of ACL injury (Kaplan, 1992; Murphy, 1992). One or both of these bone bruises may be present. The bone bruises often occur when valgus forces tear the ACL and cause anterior tibial translation. Translation is greatest laterally, causing relative external rotation of the femur relative to the fixed tibia. This pivot shift allows the lateral femoral condyle to impact the posterolateral tibial plateau (Kaplan, 1992; Mink, 1988). The lateral femoral condyle bone bruise is usually near the anterior horn lateral meniscus but may be more posteriorly located if the injury occurs during knee flexion. With more severe injury, osteochondral fractures may be present with these bone bruises. MRI demonstrates a linear, hypointense subchondral fracture line or cortical contour alterations. Lateral radiographs may reveal a corresponding "deep lateral femoral-notch sign" that manifests as an exaggerated (>1.5 mm-deep) condylopatellar notch of the lateral femoral condyle (Cobby, 1992; Kezdi-Rogus, 1994). A fracture of the posterior aspect of the lateral tibial plateau may be observed on lateral radiographs; this correlates with the posterolateral tibial pivot-shift bone bruise. This is visualized as a subtle impaction fracture or as a posterior capsular bony avulsion fragment (Stallenberg, 1993). Characteristic pivot-shift bone bruises (and osteochondral fractures) of the tibia or femur indicate a greater than 90% likelihood of ACL injury (Stoller, 1997). However, pivot-shift bone bruises without ACL tears occur more often in the pediatric population (Snearly, 1996). Contrecoup bone bruises occur frequently with ACL tears and involve the posteromedial tibial plateau at, or near, the semimembranosus tendon insertion. When a contrecoup bone bruise is present, the incidence of meniscal tears, overt or occult, is exceptionally high (Chan, 1999; Kaplan, 1999). Bone bruises on MRIs were originally reported to resolve within about 6 weeks (Gentili, 1994). However, a study demonstrated persistent visible bone bruises on MRIs in all patients at 12-14 weeks (Davies, 2004). MRI-enabled diagnosis of pivot-shift bone bruises and osteochondral fractures probably has significant prognostic implications. A subset of patients with subchondral bone bruises have long-term sequelae (osteoarthritis) related to overlying chondral injury or subsequent subchondral osseous collapse. In fact, investigators increasingly suspect that these injuries may be a principle determinant of the natural history of patients with ACL tears, ie, the bone bruises are a marker for underlying cartilage at risk (personal communication). Further long-term studies are needed to assess this issue. Anterior translocation of the tibia This secondary sign is related to the anterior drawer sign of instability elicited on physical examination. The radiologist should seek this finding on a sagittal image through the middle of the lateral femoral condyle. If the tibia translocates anteriorly to the extent that the distance between vertical tangent lines through the posterior margins of the femur and tibia exceeds 5 mm, acute or chronic ACL tear is highly likely (Vahey, 1993; Chan, 1994). Tibial translocation is also present if a vertical line tangent to the posterior cortex of the tibial plateau courses through, or anterior to, the posterior horn meniscus (the uncovered meniscus sign). This occurs because the meniscus remains with the femur as the tibia translates anteriorly. Fracture of the tibial spine The ACL does not actually insert on the anterior tibial spine; it inserts immediately lateral to it. Nevertheless, the possibility of ACL insufficiency or a concurrent ACL tear should be borne in mind when an intercondylar eminence fracture is detected. Tibial spine avulsion with ACL insufficiency or injury indicates a hyperextension mechanism in most cases, and it is relatively more common in the pediatric population (only 5% of adults with an ACL injury and a higher percentage in children). These injuries are often isolated in children, but they imply a high-force injury in adults and other internal derangements are usually present (Rogers, 1982; Kendal, 1992; Stoller, 1997; Toye, 2002). Several tibial spine fracture-classification systems have been proposed (Oostvogel, 1988). Treatment is somewhat controversial; however, surgical intervention is most likely with displaced larger fractures. Segond fracture A Segond fracture is a stereotypical fracture of the tibia that has a 75-100% association with ACL tear (Resnick, 1995). The Segond fracture is an elliptical, vertical, 3 x 10-mm bone fragment paralleling the lateral tibial cortex about 4 mm distal to the plateau. The fragment is best seen on a true anteroposterior view (Resnick, 1995) or a tunnel view (Goldman, 1998). It should be distinguished from a Gerdy tubercle bony avulsion (with iliotibial band stress) anteriorly, which is optimally seen on a radiograph with external rotation (Resnick, 1995). Segond fractures have been historically attributed to avulsion of the middle third of the meniscotibial capsular ligament, but a report implicates contributions of the iliotibial band and LCL as well (Campos, 2001). In the acute setting, MRI often shows a bone bruise of the adjacent edge of lateral tibial plateau secondary to meniscotibial ligament avulsion. The adjacent Segond fragment may be difficult to visualize (Weber, 1991). If observed, the bone fragment demonstrates a marrow-edema pattern; long term, it usually shows isointensity relative to marrow and may fuse to the underlying bone (Resnick, 1995). The tibial avulsion, Segond, and posterolateral tibial plateau fractures described above may be difficult to detect on both radiographs and MRI (Delzell, 1996; Weber, 1991). The radiologist should carefully scrutinize these areas (including tibia at insertion of lateral capsule adjacent to Segond fracture) for telltale bone bruises or subtle cortical alterations on MRIs in patients with clinically suspected ACL tear but no obvious MRI findings. Radiographic (Kezdi-Rogus, 1994) or CT correlation may be helpful in difficult cases. Less useful secondary signs of ACL tear Several secondary signs of ACL injury have relatively low specificity for ACL tear and are less useful than the signs discussed above. Buckling or redundancy of the PCL (Boeree, 1992) occurs frequently with ACL tears, but it also occurs with hyperextension of the normal knee (Gentili, 1994) and with quadriceps dysfunction (Stoller, 1997). Edema in the region of the ACL is an abnormal but nonspecific finding (Vahey, 1991); other evidence is needed to make a definitive diagnosis of tear. Kissing bone bruises indicate a hyperextension injury and were found to be associated with ACL tears in about 50% of patients in one study (Terzidis, 2004). Cadaveric studies have shown no normal synovial recesses in the triangular space inferior to the intersecting ACL and PCL as observed on sagittal images. Therefore, Lee et al (1996) hypothesized that fluid in this location may indirectly indicate abnormality of the cruciate ligaments, but this has not been confirmed in a clinical setting. As noted previously, fluid recesses anterior to the ACL are a common finding in normal ACLs. Avulsion fracture of the proximal fibula (termed the arcuate sign) is not considered a very reliable secondary sign of ACL tear, though it was associated with ACL tear in 13 of 18 patients in one study (Juhng, 2002). This fracture has a higher statistical association marker with varus and/or hyperextension injury to the posterolateral corner structures including the fibular collateral ligament. The likelihood of an ACL injury is markedly increased in the setting of combined PCL and LCL tears, as frequently occurs with knee dislocation. The ACL is torn in most knee dislocations (Yu, 1995). Partial ACL tear Partial tears of the ACL are common, accounting for 10-43% of all ACL tears (Fruensgaard, 1989; Noyes, 1989; Roychowdhury, 1997; Lee, 1998), and account for a higher percentage of ACL tears in the pediatric population (Prince, 2004). The natural history and optimal treatment of these injuries is still being worked out (Noyes, 1989; Umans, 1995). A tear involving less than 25% of the ACL has a favorable prognosis; a tear involving 0.5-0.75 of the ACL has a 50-86% probability of progressing to a complete tear (Noyes, 1989). Prognosis is guarded overall, with 38-62% of stable conservatively treated patients progressing to instability (Sandberg, 1987; Fruensguaard, 1989; Umans, 1995). On physical examination, partial tears are often difficult to diagnose. In cadavers, laxity is absent on physical examination and arthrometric testing when only the anteromedial band of the ACL is transected (Lintner, 1995). In the clinical setting, factors including soft tissue swelling and guarding may compromise the examiner's ability to detect mild laxity (Noyes, 1989). In the converse, some patients with mild detectable laxity on physical examination (suggestive of partial tear) actually have a complete tear (Lintner, 1995), or their injury progresses to complete tear over time (Noyes, 1989). Although MRI is highly accurate in differentiating the normal from abnormal ACL, it has lower accuracy in the diagnosis of partial tears (Gentili, 1994; Umans, 1995; Lawrance, 1996). Reliably differentiating high-grade partial ACL tears from complete tears is not always possible. For example, even patients with a completely nonvisualized ACL in association with hemorrhage, or patients with an abnormally horizontal ACL axis on MRI, occasionally have partial rather than complete tears (Umans, 1995). Further, secondary signs of ACL injury, such as bone bruises, while certainly correlating generally with severity of injury, do not necessarily indicate a complete tear (McCauley, 1994). In the converse, patients with normal knee MRIs are found very occasionally to have partial ACL tears. Underinterpretation can also occur when patients with only subtle abnormalities suggesting a partial injury (eg, mild acute angulation of an otherwise normal ACL) in fact may have a complete tear. False-negative MRIs in patients with a partial ACL tear are especially likely in the nonacute setting. These limitations notwithstanding, MRI clearly does allow diagnosis of some partial tears missed on physical examination; positive MRI findings should not be ignored outright on account of a negative result on the Lachman test. MRI has at least a theoretical advantage over the criterion standards of arthroscopy or arthrotomy with regard to partial intrasubstance tears. Also, several encouraging studies have indicated that MRI findings can stratify patients into high risk (complete vs high-grade partial) or low risk (low-grade partial vs negative) groups (Zeiss, 1995; Roychowdhury, 1997 and 1998; Rubin, 1998; Chen, 2002). As such, MRI maintains utility in deciding between surgical and non-surgical treatment. In view of the information presented previously, what should be the approach of the MRI reader in interpreting the ACL? When an obvious acute ACL tear is noted with direct (and possibly indirect) signs of injury, the reader should describe an acute high-grade tear. A partial tear can be specifically suggested in the appropriate clinical setting when unequivocal direct signs of ACL tear are present but at least one section shows a straight, taut-appearing ACL. A partial tear can be suggested when only partial-thickness increased signal intensity is observed in the substance of the ACL (Chen, 2002). Care must be taken to avoid mistaking this finding for artifactual partial-volume related imaging of signal related to the normal synovial recesses that closely invest the ACL (Umans, 1995). When positive secondary signs of ACL tear (eg, pivot shift bone bruises) are present but the ACL proper appears normal, or when only equivocal primary MR signs of ACL tear are found and secondary signs are absent, the reader may suggest the possibility of ACL injury (low-grade tear versus a normal ACL). Such problem cases obviously need clinical and Lachman-test correlation, and the MRI reader could consider additional problem-solving sequences, such as those discussed in the techniques section. Vincken et al (2002) reported that the decreased accuracy of MRI for the diagnosis of partial ACL tear is to some degree irrelevant to patient care. This is because high-grade tears usually coexist with other major knee injuries evident on MRIs, which cause patients to be appropriately selected for endoscopy. In the converse, relatively low-grade tears (which may benefit from a trial of conservative treatment) are less likely to have MRI diagnoses of comorbid knee injuries that promote endoscopy. In other words, the high composite diagnostic accuracy of MRI of the knee probably still directs appropriate treatment in most patients. Treatment recommendations for patients with partial ACL tears are evolving. Factors favoring conservative treatment include advanced age, a normal or near-normal Lachman result, low athletic demands, and less than 50% involvement of the ACL fibers on arthroscopy. Most young and highly active patients, patients with a clearly abnormal Lachman result, and patients with greater than 50% or posterolateral band involvement on arthroscopy are best treated with ACL reconstruction (Kocher, 2002; Fujimoto, 2002). CHRONIC ACL TEARS AND MISCELLANEOUS CONDITIONSSection 8 of 11
Chronic ACL tear The MR reader will not uncommonly encounter non-acute ACL tears. These injuries are often associated with medial meniscal tears (Allen, 2000), and secondary osteoarthritis. MRI signs of chronic ACL tear are largely the same as those of acute ACL injury except that bone bruises and edema about the knee are absent (Vahey, 1991; Dimon, 1998). The most common MR finding with chronic tear is a fragmented ACL (Vahey, 1991). Complete nonvisualization of the ACL may also occur with only fat signal intensity evident in the lateral intercondylar notch, the "empty notch" sign (Remer, 1992). The chronically torn ACL tear may attach to the PCL (Stoller, 1997). The authors, however, have noted that this is most often an endoscopic observation and is less frequently appreciated on MRI, even in retrospect. Patients with this finding may have a clinical endpoint of anterior translation of the tibia with Lachman testing, resulting in a false-negative finding on clinical examination. While the diagnosis of chronic ACL tear by MRI is usually straightforward. However, in some instances, the only sign of a chronic ACL tear is a subtle angulated appearance or flattened axis of the ACL (Moore, 2002), and, in fact, a chronic nondisplaced ACL tear may appear entirely normal (Vahey, 1991). This presumably occurs because hypointense mature scarring may be difficult to distinguish from the normal low-signal ligament. Therefore, the MR reader must be especially diligent in the nonacute setting to avoid underdiagnosing chronic tears (Vahey, 1993). In the setting of a positive Lachman test and a suspected chronic tear, a negative MRI should be viewed as a possible false negative. Mucoid degeneration and ganglion cysts of the ACL Mucoid degeneration of the ACL can mimic ACL tear (McIntyre, 2001; Narvekar, 2004). The etiology of this uncommon entity is uncertain, but it may lie along a continuum of senescent degeneration of the ligament (McIntyre, 2001). Reported patients have generally been older than 30 years. Patients may be asymptomatic, but they frequently have pain and limited flexion of the knee. The knee is stable, with a negative result on the Lachman test. On arthroscopy, the ACL is enlarged and often impinges on the intercondylar notch sidewalls or roof. Histologic examination of the ligament demonstrates extensive, patchy, yellow internal deposits, which represent a mixture of fibrous elements and mucoid degeneration. Treatment has consisted of mainly meticulous piece-by-piece debulking of the yellowish material. Some fascicles of the ACL may be unavoidably sacrificed in this procedure. Notchplasty may also be performed to reduce ACL-notch impingement. MRI appearances are characteristic. The ACL is enlarged with diffusely increased non–fluidlike increased signal intensity. The still-visible linear ACL fascicles often give the ACL a striated celery-stalk appearance. Appearances are readily mistaken for those of an interstitial ACL tear; however, a discordant history, a negative Lachman result, and a lack of secondary signs of ACL tear usually indicate the correct diagnosis (Stoller, 1997; McIntyre, 2000; Fealy, 2001; Bergin, 2004; Narvecar, 2004; Nishimori, 2004). Intraligamentous and extraligamentous ganglion cysts of the ACL are a related but distinct entity. The extraligamentous cysts are extremely common but do not present a diagnostic problem. These appear as well-defined lobular, often septated, cysts immediately adjacent to the ACL. These cysts are usually asymptomatic incidental findings, though a variety of symptoms have been reported. The smaller cysts may be difficult to differentiate from normal synovial recesses (Bergin, 2004). Ganglion cysts in the substance of the ACL are less common but have been reported in all age groups (Recht, 1994; Kang, 1995; Do-Dai, 1996; Jordan, 1999; Kumar, 1999). On MRI, these appear as a fusiform well-defined cysts oriented along the long axis of the substance of an otherwise normal-appearing ACL. When the cysts are small, this appearance may be mistaken for that of a partial ACL tear (Jordan, 1999), but differentiation is usually not difficult. Several authors have described ACL ganglion cysts in young patients that developed after trauma (Do-Dai, 1996; Kakutani, 2003). As in diffuse mucoid degeneration of the ACL described above, the knee is stable with a negative Lachman result. However, symptoms may be clinically significant, and patients often benefit from arthroscopic probing with release of the mucinous material with or without partial debridement (Do-Dai, 1996; Bergin, 2004). Ganglion cysts of the ACL can be difficult to appreciate on standard anterior portal arthroscopy. Therefore, diagnosis may depend on MRI; the MRI reader who recognizes the abnormality may alert the arthroscopist to probe the ACL or add a posterior portal approach (Fealy, 2001; McIntyre, 2001; Bergin, 2002). Deltoid-shaped ACL Calpur et al (2004) reported 2 patients with a markedly widened deltoid (triangular) distal ACL at the tibial insertion. Symptomatic impingement of the intercondylar-notch structures was reported, and successful trimming of the distal ACL (ligamentoplasty) was performed in both patients. SUMMARYSection 9 of 11
Recommendations The goal of this article is to educate the MR reader in interpretation of the normal and abnormal ACL. To this end, several points should be reemphasized. Regarding generation of images, one should obtain spin-echo or fat-saturated fast spin-echo images in all 3 planes, including both T1- and T2-weighted sagittal images. The authors recommend the technique of prescribing sagittal images no more than 10° oblique to a perpendicular to the bicondylar line on an axial scout image. On sagittal images, one should critically evaluate the ACL axis relative to the intercondylar roof. The proximal and distal aspect of the ACL should also be carefully evaluated; tears or osseous avulsions may be readily overlooked in these locations. Axial sequences are especially useful in evaluating the proximal ACL. One should become familiar with normal and abnormal appearances of the ACL in all planes. One should also look for secondary signs of ACL tear, including subtle focal marrow edema of the tibial plateau observed with tibial spine avulsion and focal bone edema of the lateral edge of the tibia observed in association with an occult Segond fracture. Radiographic correlation may be helpful in these settings. One should be wary of a normal MRI in a patient with a positive Lachman test and a history of remote injury; look for subtle signs such as angulation or waviness of the ACL as perhaps the only clue of a chronic nondisplaced tear. In the converse, be wary of overdiagnosing an interstitial tear in patients with mucoid degeneration of the ACL. Lachman test results should help in differentiating these entities. In difficult cases, obtain additional history. One may also consider performing an additional double-oblique oblique coronal thin-section T2-weighted sequence aligned along the long axis of the ACL. Finally, one should avoid satisfaction-of-search error. When the ACL is torn, look especially diligently for other internal derangements. Detection of meniscal tears is more difficult in the setting of an acute ACL tear than in other settings. Future of ACL imaging Much information still needs to be learned about MRI of the ACL. Studies to determine the association between MRI appearances of the ACL and long-term functional patient outcomes are lacking (Boks, 2006). The role of articular cartilaginous injury in the natural history of patients with ACL tear has yet to be elucidated. In addition, room for improvement in MRI diagnosis of partial tears of the ACL is needed. Fortunately, faster and better MR images can be anticipated as a result of continued technological advances in instrumentation, software, and contrast agents. Thank you to Debra Blaylock and Preston Library staff for their assistance with this article. MULTIMEDIASection 10 of 11
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
