Anatomy of the Anterior Cruciate Ligament with Regard to Its Two Bundles

Abstract

The anterior cruciate ligament (ACL) consists of two major fiber bundles, namely the anteromedial and posterolateral bundle. When the knee is extended, the posterolateral bundle (PL) is tight and the anteromedial (AM) bundle is moderately lax. As the knee is flexed, the femoral attachment of the ACL becomes a more horizontal orientation; causing the AM bundle to tighten and the PL bundle to relax. There is some degree of variability for the femoral origin of the anterome-dial and posterolateral bundle. The anteromedial bundle is located proximal and anterior in the femoral ACL origin (high and deep in the notch when the knee is flexed at 90°); the posterolateral bundle starts in the distal and posterior aspect of the femoral ACL origin (shallow and low when the knee is flexed at 90°). In the frontal plane the anteromedial bundle origin is in the 10:30 clock position and the postero-lateral bundle origin in the 9:30 clock position. At the tibial insertion the ACL fans out to form the foot region. The anteromedial bundle insertion is in the anterior part of the tibial ACL footprint, the posterolateral bundle in the posterior part. While the anteromedial bundle is the primary restraint against anterior tibial translation, the posterolateral bundle tends to stabilize the knee near full extension, particularly against rotatory loads.

One of the first anatomical descriptions of the anterior cruciate ligament (ACL) can be found written on an Egyptian papyrus scroll dating back to 3000 BC.¹⁰ Hippocrates (460-370 BC) described a subluxation of the human knee related to ACL injury. However, the ligament was named by Claudius Galen of Pergamon as “ligamenta genu cruciate” (129-199 BC).¹⁰ The ACL has been one of the most frequently studied structures of the musculoskeletal system during the last decades. In 2005, an Internet-based literature review of the National Library of Medicine catalogue for the keyword “anterior cruciate ligament” resulted in 6383 hits, reflecting its importance in basic and clinical research. It has been studied from many points of view, including biomechanical behavior of the intact ligament and replacement grafts, mechanisms of failure, treatment, surgical techniques, and postoperative rehabilitation protocols.

For many surgeons the most fundamental aims of reconstruction are to restore function and prevent late OA.^16-18,24,33 Some authors believe current ACL reconstruction techniques using tendon grafts anchored in one femoral and one tibial tunnel only partly achieve this goal. ^{8,11,13,17,19,21,34,36,38,45,46,55,58-60}

The ACL is not just a band of dense connective tissue connecting the femur and the tibia. For many years it has been well recognized the ACL consists of two bundles, the anteromedial and posterolateral bundle, so named for the orientation of their tibial insertions (Fig 1). ^{2-4,6,9,22,30,31,37,40,43,44,56,58,60,64,65} Even though anatomical dissections suggest there may be even more bundles, most believe there are only two major functional bundles.^{2-4,6,7,14,22,30,31,37,40,43,44,56,58,60,64,65} Recent bio-mechanical studies of quadruple-loop semitendinosus and gracilis tendon grafts and bone-patellar tendon-bone grafts suggest these reconstructions were successful in limiting anterior tibial translation in response to an anterior tibial load, but insufficient to control tibial rotation.^{20,21,57,58,62} In an in vivo study Georgoulis et al^20,21 showed single bundle reconstruction restores anterior tibial translation, whereas rotation remains increased during high demand activities. ACL reconstruction is potentially limiting since it designed primarily to replicate the anatomy of only the anteromedial bundle. Based upon the relative amount of literature, we believe the other functional bundle, the posterolateral, has not received sufficient attention.

Fig 1: This figure shows a right knee from the anterior aspect. The patellar tendon and the surrounding soft tissue have been removed to inspect the ACL. Note the two distinct bundles, anteromedial (AM) and posterolateral (PL).

Single bundle ACL reconstruction is a common procedure to restore knee stability, and good-to-excellent clinical results have been reported.^16,61 However, critical reviews of the literature reveal the success rate after ACL reconstruction reported for single-bundle ACL reconstruction is not always satisfying.^16,61 Lohmander et al³³ showed a high incidence of radiological signs of osteoarthritis at 12 years after single bundle ACL reconstruction. Therefore, a more anatomical approach to reconstructing both bundles of the ACL has evoked the interest of many surgeons. ^{1,8,11,13,17,34,36,38,45,51,55,59,60}

To achieve satisfactory surgical outcome after ACL reconstruction, basic knowledge of the anatomy of the ACL is essential. We therefore describe the anatomy of the native ACL with regard to its two bundles.

Development of the ACL

The knee originates from vascular femoral and tibial mesenchyme in the fourth week of gestation between the blastema of femur and tibia.^41,42,64,65 At this time, the fibroblasts in the ligament start to align to the axis of strain of the ACL. By 9 weeks, the cruciate ligaments are composed of numerous immature fibroblasts having scanty cytoplasm and fusiform nuclei (Fig 2).^42,43 These fibroblasts produce the extracellular matrix of the ACL. After week 20, the remaining development consists of marked growth with little change in form. Tena-Arregui et al⁵⁴ performed arthroscopy on the knee of fetuses with a gestational age of 24 to 40 weeks. At these stages two main bundles were already detectable, but the bundles seemed more parallel when compared to the bundle orientation of the adult ACL.⁵⁴

Fig 2: This figure shows a fetal knee in the 10^th week after gestation. Note the early organization of the ACL between femur (F) and tibia (T).

The early manifestation of the ACL with two different bundles in the fetal knee suggests early development of the knee joint is guided by the ACL. That cruciate ligaments are present at this early stage of development could lead to the assumption they interact with the resulting shape of the femoral condyles and the tibial plateau.³³

Anatomy of the Intercondylar Fossa

The intercondylar fossa is the groove between both femoral condyles and the origin for the cruciate ligaments (Fig 3). Therefore, knowing topographical anatomy of this region is essential for surgeons who perform arthroscopy and cruciate ligament surgery.

Fig 3A: B. (A) This figure shows a lateral femoral condyle. The anteromedial bundle is located high and deep in the intercondylar fossa, whereas the postero-lateral bundle is located more low and shallow. Because arthroscopic surgery is normally performed with the knee in 90° of flexion, in the clinical literature the notch roof is considered high (anatomic description, anterior), the anterior rim is considered shallow, the posterior part is considered deep, and the posterior part is considered low. (B) The posterior end of the intercondylar fossa is marked by the intercondylar line (arrows). This line borders on the Facies poplitea (FP) of the distal femur. The AM origin is located at the transition of the intercondylar line to the articular cartilage.

The anterior outlet of the intercondylar fossa is shaped like a Gothic arch with a small notch for the ACL.^4-6,32,47 The posterior edge of the intercondylar fossa is marked by the intercondylar line (Linea intercondylaris).^42,43 This line borders to the Facies poplitea of the distal femur (Fig 3).^42,43 There is a discrepancy between the anatomical and clinical literature when describing the topographical anatomy of the intercondylar fossa. In the anatomical literature, the orientations in the intercondylar fossa are determined as proximal, distal, anterior, or posterior (Fig 3).⁶⁴ However, ACL surgery is normally performed with the knee in 90° of flexion. Therefore in the clinical literature the directions in the intercondylar fossa are determined as high, low, shallow, and deep (Fig 3).⁶⁴

The intercondylar fossa has its maximum diameter in the posterior part and converges towards the anterior direction. The width of the notch is apparently smaller in women when compared to men.^5,32,47 This could probably be an additional factor explaining the higher incidence of ACL injuries in female athletes besides gender-specific neuromuscular differences. The notch width index is used for the ratio epicondylar width to notch width.^5,32,47,49 Because patients with bilateral ACL rupture showed a smaller notch width index, the risk of sustaining an ACL rupture could be higher.^5,32,47,49

On conventional lateral knee radiographs the notch roof projects as a dense profile known as Blumensaat's line.⁵ The average angle between the longitudinal axis of the femur and the notch roof (notch roof angle) is given as 37°, but varies between 23° and 60°.^4,28,62 Knees with a small notch roof angle are also known as nonforgiving knees.²⁸ In ACL reconstruction in these knees, the position of the tibial bone tunnel is rather critical and an anteriorly misplaced tunnel should be avoided to prevent graft impingement by the anterior rim of the intercondylar fossa.²⁸

In intact knees without any signs of arthritic changes the size of the ACL matches that of the intercondylar fossa. But it is well known substantial osteophyte formation and stenosis of the outlet of the anterior outlet of the intercondylar notch occurs early in the ACL deficient knee.⁴³ In these cases, the ACL has apparently inadequate space. Based upon this concept, some authors recommend notchplasty during ACL reconstruction.^18,41,48

Gross Anatomy of the ACL

The ACL is enveloped into the synovial membrane of the human knee, which by definition places the ligament intraarticular but extrasynovial.⁴³ However, from a surgical standpoint, the ACL must be regarded only as an intraarticular structure because it can be reconstructed only intraarticularly and cannot be exteriorized from the articular space.⁴³

The ligament originates at the medial side of the lateral femoral condyle and runs an oblique course through the intercondylar fossa distoanteromedial to the insertion at the medial tibial eminence. The axis of the long diameter of the ACL is tilted 26° ± 6° forward from the vertical.^43,64 When the knee is flexed, as during surgery, the ligament seems to turn itself in a lateral spiral. This external rotation is approximately 90° as the fibers approach the tibial surface. The twist of the fibers of the ACL is a result of the orientation of its bony attachments. The femoral attachment is oriented primarily in the longitudinal axis of the femur, whereas the tibial attachment is in the anteroposterior axis of the tibia.⁴³

The narrowest diameter of the ACL occurs in the mid-substance (Figs 1, 4, 5).²³ Anderson et al¹ reported an oval shape with an area of 36 mm² and 44 mm² for females and males, respectively. Harner et al²³ digitized the insertion areas of the ACL and reported, relative to the ligament midsubstance, the tibial and femoral insertion of the ACL was over 3.5 times larger. The broad femoral and tibial insertion area with respect to the narrow diameter of the midsubstance makes selection of tunnel site placement challenging because of the limited size of single-bundle grafts.¹⁸

Fig 4A: B. (A) This figure shows the human knee with the medial femoral condyle removed. This allows inspection of the femoral attachment of the ACL. Fibers of the ACL run approximately parallel close to the extended position. In extension the fibers of the PL bundle are tensed. Note the extension of the fibers from the intercondylar line all the way to the cartilage-margin. Because of the anatomical shape of the insertion areas at the femoral and tibial attachment site, anterior fibers of the ACL are able to bend around the anterior edge of the intercondylar notch (arrow). (B) In flexion the PL origin rotates anteriorly and the AM bundle becomes tensed.

Fig 5: This figure shows that in extension the anterior ACL fibers tuck under the anterior roof of the intercondylar fossa (AM = anteromedial bundle, PL = posterolateral bundle). Note the proximity of the AM bundle insertion to the lateral meniscus (LM).

At the tibial side, the fiber bundles of the ACL fan out and form a so-called foot region (Fig 5).⁶ This allows the ACL to tuck under the roof of the intercondylar notch (Fig 5).^6,33,49 In full extension, the anterior fibers of the ACL course around the anterior edge of the intercondylar notch (Figs 4, 5). This wrapping around is called physiological impingement.^42,43 This specific anatomy causes concern for ACL reconstruction because common grafts do not possess such a wrapping region. If a straight graft is inserted in the anterior part of the tibial ACL it tends to impinge with the notch in slight degrees of flexion (Fig 4).⁴² Notch impingement due to anterior tibial tunnel position was considered a common cause of postoperative deficits in range of motion, resulting in a lack of extension. To prevent notch impingement, most surgeons tend to place the tunnel in the posterior part of the tibial insertion (PL bundle insertion).^28,35,51

The Anteromedial and Posterolateral Bundle

It has long been realized the ACL does not function as a simple band of fibers with constant tension as the knee moves. However, the differentiation of the ACL into different bundles is controversial.^{2-4,6,7,22,30,31,37,40,43,44,56,58,60,64,65} Odensten and Gilquist⁴⁰ examined the ACL histologically and found no evidence to separate the ligament into two bundles. Amis and Dawkins³ divided the ACL into anteromedial, intermediate, and posterolateral bundles. Even though there is disagreement on the actual anatomic division of the ACL−there is no histological separation of different bundles whether there are two, three, or multiple functional bundles−there is general consensus two functional bands can be distinguished as the tension varies among the fibers in the ligament with range of motion. ^{2,3,6,8,19,22,31,43,58,59} The differentiation of the ACL into two functional bundles, the anteromedial and posterolateral bundle, seems an oversimplification, but the two bundle description of the fibers of the ACL has widely been accepted (Fig 1).

The terminology of the bundles is determined according to their tibial insertion, with the fibers of the anteromedial bundle originating in the most proximal part of the femoral insertion and inserting at the anteromedial tibial insertion.^22,56 Fibers of the posterolateral bundle originate distally at the femur and insert on the posterolateral part of the tibial insertion.^{22,31,41,62,64} In the frontal plane, the anteromedial bundle has a more vertical orientation (approximately 70° to the knee base line) while the posterolateral bundle is oriented more horizontally (approximately 55° to the knee base line).^43,64

Anatomical and in vivo studies from our laboratory suggest a great variance in size of the two bundles.^43,64 Length and diameter of the native ACL play an important role for choosing the type of graft and for the preparation of the graft in ACL reconstruction. Especially for some extracortical fixation techniques, the intraarticular length of the native ACL may be important because it will determine the length of the graft available in the femoral and tibial tunnel. Length and orientation of the fibers of the ACL change throughout the passive flexion and extension and tibial internal and external rotation.^3,15,26,27 The length of the ACL fibers range from 22 mm to 41 mm with a mean of 32 mm.^31,43 However, these measurements have been made for fibers of the ACL belonging possibly to the anteromedial fiber bundle. Kummer and Yamamoto³¹ measured the intraarticular length of the posterolateral bundle in 50 cadavers and reported a length of 17.8 mm.

For ACL reconstruction and femoral tunnel placement it might be important, as Hefzy and Grood²⁶ demonstrated, fiber length is affected more by varying the femoral attachment. Moving the tibial location had only a small effect; however, anteroposterior and proximodistal variations at the femoral insertion site had strong effects on length patterns.²⁶

Function of the Anteromedial and Posterolateral Bundle

It is generally accepted the anterior cruciate ligament is the primary restraint to anterior tibial translation. However, the lack of congruency because of the concave and convex shape of the tibiofemoral joint also allows tibial internal and external rotations.² The lateral compartment is more important in tibial rotation laxity because it is more mobile.² Therefore, the anterior tibial drawer is normally accompanied by a coupled tibial rotation. In the ACL-deficient knee the axis of rotation shifts more medially and the tibial rotation causes a coupled anterior tibial translation magnifying the movements of the tibial plateau.² This raises the question of whether the ACL also plays a role for stabilization against rotatory loads. Cadaveric studies revealed single-bundle ACL reconstruction procedures are: (1) successful in limiting anterior tibial translation in response to an anterior tibial load; (2) insufficient to control a combined rotatory load of internal and valgus torque; and (3) stress the function of the posterolateral bundle.^19,58,65

When the knee is extended, the posterolateral bundle is tight (Fig 4) and the anteromedial bundle is moderately lax. As the knee is flexed, the femoral attachment of the ACL moves to a more horizontal orientation, causing the anteromedial bundle to tighten and the posterolateral bundle to loosen up.³ However, none of the ACL fibers behave isometrically.³

The anteromedial and the posterolateral bundle also have independent functions in providing stability to the knee.^19,58,65 Amis and Dawkins³ showed the contribution of the PL bundle to resisting anterior draw in extension, and the AM bundle becoming dominant at 90° of flexion. Recent studies have suggested a reciprocal relationship exists in the force between the two bundles of the ACL.^19,63 In response to an anterotibial load, the in situ force in the posterolateral bundle was highest at full extension and decreased with increasing flexion.^19,63 The in situ force in the anteromedial bundle was lower than in the posterolateral bundle at full extension, but increased with increasing flexion, reaching a maximum at 60° of flexion.¹⁹ Under a combined rotatory load, the in situ force of the posterolateral bundle was higher at 15° and lower at 30° of flexion.¹⁹ The in situ force in the anteromedial bundle was similar at 15° and 30° of knee flexion.¹⁹ Anterior translation under a combined rotatory load (valgus and internal rotation) is a biomechanical model to simulate the pivot shift.⁵⁸ Our study⁶³ suggested knees with a PL bundle transsection had substantially higher anterior tibial translation at 0 and 30° of knee flexion when compared to the intact knee. These findings demonstrate the importance of the posterolateral bundle for stabilizing knee, especially when the knee is near full extension. To reproduce the in situ forces of the normal ACL, reconstruction techniques should take into account the role of the posterolateral bundle in addition to the anteromedial bundle.^19,63

Consequently, Yagi et al⁵⁸ investigated knee kinematics and in situ force of a single-bundle and a double-bundle ACL reconstruction technique (one tibial and two femoral tunnels). In their study, anterotibial translation for the double-bundle reconstruction was substantially closer to the intact knee than was the single-bundle reconstruction.⁵⁸ The in situ force normalized to the intact ACL for the anatomic reconstruction was 97% ± 9%, whereas the single-bundle reconstruction was only 89% ± 13%.⁵⁸ With a combined rotatory load, the normalized in situ force for the single-bundle and anatomic reconstructions at 30° of flexion was 66% ± 40% and 91% ± 35%, respectively. In this biomechanical study, anatomic reconstruction produced a better biomechanical outcome, especially during rotatory loads.⁵⁸

Origin and Insertion of the Anteromedial and Posterolateral Bundle

Misplacement of the bone tunnels may be one of the most frequent causes of ACL revision surgery.^7,24,39 For anatomic ACL reconstruction this means when creating two femoral and two tibial tunnels instead of one femoral and one tibial tunnel, the risk of tunnel misplacement is high.

It is essential to know the anatomy of the femoral origin and the tibial insertion for placement of bone tunnels in ACL reconstruction. Earlier studies have mainly focused on the origin of the anteromedial bundle. However, with rising interest in anatomical ACL reconstruction techniques, descriptions of the distinct origin and insertion of the two functional bundles are needed.

Femoral Origin

The femoral origin of the ACL at the posterior part of the medial surface of the lateral femoral condyle (Figs 3, 4) has been described in several studies.^{22,23,40,43,64} Girgis et al²² described the femoral origin as a segment of a circle with its anterior side straight and the posterior side convex. According to these authors, there is a distance of 2 to 3 mm between the posterior border of the ACL origin and the articular cartilage of the femur.²² This description is not in accordance with our anatomical studies,^43,64 which suggest the femoral origin of the ACL is more oval and more posterior (Figs 3, 4).^43,64 Other authors using laser-digitizing methods also found the ACL origin more circular²³ or more oval⁴⁰ with a mean length of 18 mm and a width of 11 mm.^23,40 A histological study suggests the ACL insertion directly borders the intercondylar line superiorly and the articular cartilage in the inferior part of the wall of the lateral condyles.^41,42 It is important to appreciate the three-dimensional shape of the ACL insertion which extends from the intercondylar roof downwards along the cartilage contour to the inferior wall of the lateral femoral condyle (Fig 4).

From a clinical viewpoint, the surgeon should be aware the ACL origin lies behind a small bony ridge, referred to as residents' ridge, which could be one reason for anterior tunnel misplacement.²⁹ It can be quite difficult to visualize the entire ACL origin via lateral arthroscopic portals, which are considered as standard portals for most arthroscopic operations (Fig 6). For better visualization of the femoral ACL origin some experts recommend switching the scope from the anterolateral standard portal to an anteromedial portal (Fig 6).^55,59,60,64

Fig 6A: E. (A) This figure shows a macerated distal femur in a view from the posteromedial aspect. The dark area marks the femoral origin of the ACL. The AM bundle is located at the transition of the intercondylar line to the cartilage margin. The distance between the center of the anteromedial bundle and the intercondylar line (dotted line) or the articular cartilage varied between 5 and 6 mm. The posterolateral bundle originates in the anterior and inferior aspect of the femoral origin. Inferiorly there is a distance of only 4 to 5 mm between the center of the posterolateral bundle and the inferior edge of the joint cartilage (dotted line). The distance between the centers of the bundles varies between 8 and 10 mm. (B) This figure shows the arthroscopic view with the arthroscope from the high anterolateral portal. (C) A better visualization of the ACL origin is possible via the anteromedial portal. (D) The figure shows an arthroscopic view of the transition of the intercondylar line to the cartilage margin. This is the region where the AM bundle is attached. (E) This is an arthroscopic picture of an anatomical specimen illustrating the transition of the intercondylar line to the cartilage margin.

In single-bundle ACL reconstruction femoral anterior tunnel misplacement is a common pitfall in ACL surgery and a common cause of revision.^7,24 Tunnel placement too anteriorly may result in instability and flexion deficit.^7,24 If the tunnel is located anterior to the center of rotation, the graft becomes stretched when the knee is flexed. This results in a flexion deficit and elongation of the graft, leading to instability.

Tunnel misplacement occurs not only in the sagittal plane but also in the frontal plane. Placement of the tunnel high in the notch results in stretching of the ACL graft over the posterior cruciate ligament (PCL).⁵² For the patient, this type of graft placement might result in an extension block.⁵² A graft placed near the center of rotation, such as the 12 o'clock position of the knee, may be unable to resist rotatory loads because of the lack of a moment arm. As a result, some surgeons advocate a more lateral femoral tunnel (even as far as the 9 o'clock position) to increase the moment arm in single-bundle reconstructions.^18,24,48

Although many authors have described the anatomical footprints of the ACL, information on the origin of the anteromedial and the posterolateral bundle remains vague. Our anatomical studies suggest some degree of variability for the femoral origin of anteromedial and posterolateral bundle, but in all specimens the anteromedial bundle was located in the proximal and anterior aspect of the femoral origin (high and deep in the notch in the flexed knee).⁴³ The distance between the center of the anteromedial bundle and the intercondylar line varied between 5 and 6 mm. In the frontal plane, this localization corresponds to a 10:30 position (1:30, respectively) (Figs 3, 4). The posterolateral bundle originates in the anterior and inferior aspect of the femoral origin. Inferiorly there is a distance of only 4 to 5 cm between the center of the posterolateral bundle and the inferior edge of the joint cartilage (Fig 3). Yasuda et al^59,60 showed in 90° of flexion the PL point can be identified as the point 5 to 8 mm anterior to the edge of the joint cartilage. The distance between the center of the AM and PL bundle may be an important value with regard to tunnel placement. Our dissections suggest the distance between the center of AM and PL bundle varies between 8 and 10 mm depending on the size of the knee.⁶⁶ Considering the o'clock system the origin of the posterolateral bundle tends to be in the 9:30 position (Fig 3,4).⁴³ However, Yasuda et al^59,60 showed the o'clock system is not the ideal tool for describing the PL bundle location.

Tibial Insertion

The ACL insertion begins approximately 10 to 14 mm behind the anterior border of the tibia and extends to the medial and lateral tibial spine (Fig 7). In most specimens the tibial insertion is triangular in shape, but in some specimens the tibial insertion can also be oval.^{22,23,40,43,64} The diameter ranges between 10 and 13 mm in the frontal plane and 15 to 19 mm in the sagittal plane depending on the size of the knee.^{22,23,40,44,64} The sagittal diameter has clinical relevance regarding anatomic double bundle reconstruction. It is important to realize the size of the bone tunnels should be adapted to the size of the knee because it is impossible to place two 8-mm bone tunnels with a bony bridge in small knee with an insertion site of only a 15 mm diameter.

Fig 7A: B. (A) This figure shows the tibial insertion of the ACL. The ACL insertion begins behind a little bony spine behind the anterior edge of the tibia and extends to the medial and lateral tibial spine. In most specimens the tibial insertion is triangular in shape. (B) The tibial attachment is bounded by articular cartilage and meniscal attachments. The center of the insertion of the anteromedial bundle (AM) is aligned to the center of the anterior horn of the lateral meniscus (dotted line). The center of the posterolateral bundle (PL) has a distance of 7 to 8 mm to the anterior edge of the PCL (dotted line). Note the posterior root of the lateral meniscus is at risk when drilling the PL tunnel.

Because of ligament fanning, the insertion site is much larger than the midsubstance and femoral attachment of the ligament.^22,23,43 Harner et al²³ reported the tibial insertion of the ACL approximately 120% of the femoral insertion site. The anterior ACL fibers spread out anteriorly forming a flattened-out fiber area, referred to as a duck's foot.⁴ For this reason some surgeons recommend the tibial tunnel should not fill the tibial ACL attachment area anteriorly.⁶ However, it is also known a graft centered posteriorly in the tibial insertion has a steeper orientation relative to the tibia plateau than the natural ACL. This means higher forces are needed to resist a given anterior drawer force.

The literature provides only limited data about the separate insertions of the anteromedial and posterolateral bundles on the femur.^3,6,22,26 According to our observations, the center of the anteromedial bundle is located in the anterior part of the tibial insertion, has a distance of 13 to 17 mm to the anterior edge of the tibia, and is aligned with the anterior horn of the lateral meniscus.^43,64 This means an anteromedial tunnel placed in this area is at high risk for graft impingement and anteromedial tunnel placement at the tibia must be considered critical. The center of the posterolateral bundle is located in the posterior part of the tibial insertion and has a distance of 20 to 25 mm to the anterior edge of the tibia. In some specimens, the posterolateral bundle can send fibers to the posterior attachment of the lateral meniscus.^43,64 Therefore care should be taken not to injure the posterior root of the lateral meniscus when drilling the PL tunnel. Although at the tibia the main concern has been the anteroposterior direction, the ACL surgeon must remember the diameter of the graft, thereby controlling mediolateral placement of the bone tunnels. From this perspective it seems relevant that, in the frontal plane, both bundles are inserted in the middle of the tibia (Fig 7).

Radiographic Anatomy

The radiological anatomy of the origin and insertion of the ACL has great clinical relevance: knowing where the attachments project on knee radiographs is required to ascertain tunnel positions after surgery and is useful intraoperatively when assessment of tunnel placement is achieved using K-wires and fluoroscopy.

During the second European Society of Sports Traumatology, Knee Surgery and Arthroscopy (ESSKA) workshop in November 1996, there was general agreement descriptions of graft placements should be given in percentages related to some anatomical dimension.⁴ This approach would allow for different sizes of knees. The same seems true for the description of the anatomical center of the anteromedial and posterolateral bundle.

For the femur it was suggested tunnel positions be described in percentage of the notch roof length in antero-posterior and proximodistal direction.^4,65-67 Using this method the anteromedial center is approximately 30 mm away from the notch roof line and approximately 20% along point zero near the over the top position. Bernard and Hertel⁹ described the quadrant method for defining the center of the femoral insertion site of the ACL on conventional lateral radiographs (Fig 8). These authors showed the center of the ACL attachment at the lateral femoral condyle can be located at 24.8% of the distance defined by the intersection of Blumensaat's line and the contour of the lateral femoral condyle on lateral radiographs at 28.5% of the height of the lateral femoral condyle from Blumensaat's line. They concluded that by dividing the intercondylar fossa into quadrants, this point can be found inferior to the most superoposterior quadrant.⁹ Our radiographic study suggests this location may reflect the center of the anteromedial bundle rather than the center of the whole ACL origin.⁶⁶ The center of the posterolateral bundle was located in the inferoposterior quadrant (Fig 9).

Fig 8: Schematic drawing showing the methods for determining the femoral and tibial attachments of the AM and PL bundle on radiographs. Femoral the measurements were performed using the technique by Bernard and Hertel, tibial analysis were performed using the technique of Stäubli and Rauschning.

Fig 9A: E. The center of the femoral and tibial attachments of the anteromedial and posterolateral bundle have been marked with small wires and radiographs of the specimen were taken. (A) The tunnel positions can be expressed in percentages of Blumensaat's line (a) and the height of the condyles. (b) According to the quadrant method described Bernard and Hertel⁹ the center of the anteromedial bundle can be found in the superoposterior quadrant, whereas the posterolateral bundle was located in the inferoposterior quadrant. (B) This is an anteroposterior view of a femoral condyle in extension. Here the center of the PL bundle is distal to the center of the AM bundle. (C) On the tunnel view a circle can be drawn onto the notch in a clockwise fashion. On this clock, the femoral anteromedial center is at the 1:30 position whereas the femoral posterolateral center is at the 2:30 position. (D) This figure shows a lateral radiograph of a proximal tibia. On this projection the ACL attachment is located between the tip of the lateral tibial spine and a small third spine on the edge of the tibial plateau. The center of the anteromedial bundle is approximately 30% of the maximal tibial diameter and the center of the posterolateral bundle is located at 44%. (E) This figure shows an AP view of a tibia. In this plane the anteromedial and posterolateral insertions are rather in the center of the tibial plateau.

To describe the ACL origin in the frontal plane, a circle can be drawn onto the notch in a clockwise fashion, as proposed by the scientific committee of ESSKA. The clock position has been criticized by basic researchers and clinicians because it cannot take into consideration the three dimensional characteristics of the insertion of the ACL on the medial surface of the lateral femoral condyle. However, it has been useful to describe tunnel positions on tunnel view radiographs, computed tomography scans, and magnetic resonance images.⁴ On radiographs using a tunnel view (Rosenberg view), the center of the femoral anteromedial bundle is at the 10:30 position, whereas the femoral posterolateral center is at the 9:30 position (Fig 9).

On lateral radiographs the tibial ACL attachment is located between the tip of the lateral tibial spine and a small third spine on the edge of the tibial plateau (Fig 9).⁵⁰ Stäubli and Rauschning⁴³ found the ACL insertion between 25% and 62% of the sagittal tibial diameter and recommended placing the tunnel for the graft at 43%. With this method, the center of the anteromedial bundle is approximately 30% of the maximal tibial diameter and the center of the posterolateral bundle is located at 44%.

Arthroscopic Anatomy

In an in vivo study evaluating the posterolateral bundle during arthroscopy, we were able to define the two bundles in all cases.⁶⁴ However, in some cases the anteromedial bundle was overlying the posterolateral bundle and the posterolateral bundle could only be seen by careful retraction of the anteromedial with a probe (Fig 10). This might be why some orthopaedic surgeons are not aware of the presence of the posterolateral bundle. When the knee is placed in the figure of 4 position, however, the posterolateral bundle can be easily identified (Fig 10).

Fig 10A: C. (A) This figure shows an arthroscopic view of the tibial ACL insertion. (B) The center of the posterolateral bundle is 7 to 8 mm anterior to the PCL. The aiming guide for the posterolateral bundle is placed with regard of the PCL for drilling of the k-wire. The k-wire in place reveals the AM tunnel. (C) The anterior horn of the lateral meniscus (LM) corresponds to the center of the anteromedial insertion.

For arthroscopic ACL reconstruction, landmarks are needed to locate the femoral ACL origin and the tibial insertion to place the bone tunnels correctly in this area. For the localization of the tibial attachment of the ACL, two different anatomical landmarks are currently used. Some surgeons tend to prefer to orient themselves at the anterior edge of the PCL insertion. One study documented the distance between this point and the tibial ACL attachment is approximately 7 mm.³⁶ It has been argued the PCL has a variable structure,⁵ and other authors have criticized the tibial tunnel placement with respect to the PCL because the surgeon could probably tend to place the tibial tunnel too far posterior (Fig 8).^43,64 Another landmark for tibial tunnel placement is the anterior horn of the lateral meniscus, which has been described as aligned with the anteromedial aspect of the ACL insertion. Our study suggests a point 7 to 8 mm from the PCL reflects the center of the posterolateral insertion, whereas the anterior horn of the lateral meniscus might reflect the center of the anteromedial insertion (Figs 5, 7, 10).^43,64

Because of the slightly convex shape of the lateral femoral condyle with the prominent resident's ridge, the arthroscopic visualization of the femoral origin of the ACL can be technically demanding. A better visualization may be possible by switching the scope to the medial portal when determining the entrance point for the femoral tunnels (Figs 6, 11). The AM tunnel should be placed at the end of the intercondylar line where the articular cartilage begins (Figs 3, 4, 6)

Fig 11A: E. (A) This figure shows an arthroscopic view of a left knee through the anterolateral portal. The PL bundle can be seen if the anteromedial bundle is retraced with a probe. (B) When the knee is placed in the figure of 4 position, the posterolateral bundle can be easily identified. (C) This figure shows a view from the anterolateral portal into the intercondylar fossa after resection of ACL remnants and drilling of the femoral tunnels. Because of the slightly convex shape of the lateral femoral condyle with the prominent resident's ridge, the arthroscopic visualization of the femoral origin of the ACL can be technically demanding. (D) A better visualization may be possible by switching the scope to the medial portal when determining the entrance point for the femoral tunnels. The anteromedial tunnel has been already drilled. Note the space shallow and low for the PL bundle tunnel. (E) Both tunnels were drilled (LFC = lateral femoral condyle, PCL = posterior cruciate ligament).

Fine Structure

Femoral origin and the tibial insertion of the ACL have the structure of a chondral apophyseal enthesis.^41,43 The ligament itself consists of dense connective tissues and it is covered by synovial membrane.^43,53 The collagen fibrils are surrounded by connective tissue, which forms multiple fascicles in the ACL.^42,43 The major collagen of the ACL is Type I collagen, the loose connective tissue consists of Type III collagen.⁴²

Interestingly, one anatomical study revealed differences in the structure of the anteromedial and posterolateral bundle (Fig 12).⁴² In the anterior part of the anteromedial bundle, the typical cell morphology is different when compared with the typical structure of the rest of the ACL. In this region, the cells do not appear elongated. In full extension, this part of the ACL is in direct contact with the intercondylar fossa.⁴² Histological sections of this area reveal typical tenocytes and chondrocyte-like cells (Fig 12). These chondroid cells even produce small amounts of the cartilage-specific Type II collagen. Because of the direct contact of the cartilage and the ligament, the appearance of chondrocytes could be explained as a functional adaptation of the ligament to compressive stress, which is caused by the physiological impingement between the ACL and the anterior rim of the intercondylar fossa.⁴²

Fig 12A: C. (A) This is a Toluidine blue stained histological section of the anterior part of the distal third of the ACL (original magnification, ×5). (B) In the part where the ACL is in contact with the fossa (“physiological impingement”), chondrocyte-like cells aligned in rows can be found (original magnification ×40). (C) In the PL bundle, there are fusiform cells aligned between the collagen fibers (original magnification ×40).

DISCUSSION

Early descriptions on the two functional bundles of the ACL date back to an anatomical study performed by the Weber brothers in 1856.⁵⁶ Since then, many anatomical and biomechanical studies have confirmed the “two bundle concept.”^{3-7,22,30,31,35,40,43,44,56,58,60,64,65} Initial clinical descriptions of separately replacing each of the two bundles in an open fashion date back to the 1980s.^12,35,37 Despite these reports most surgeons still preferred single bundle reconstruction in the following years. Therefore, most previous anatomical studies described the whole ACL rather than the AM and PL bundles separately.^14,23,40 This is a major limitation regarding guidelines for tunnel placement of double bundle reconstruction. After several advances regarding arthroscopic technique, graft choice, and graft fixation, the double bundle reconstruction of the ACL has evoked new interest among arthroscopic surgeons in recent years.^{1,8,11,13,17,34,37,38,45,51,55,59,60} Although all of these reports describe techniques for “double bundle ACL reconstruction,” major differences between these techniques can be found. Most disagreement concerns the PL bundle location and PL tunnel drilling. Marccachi et al³⁴ recommend placing the PL bundle in the “over the top position.” From an anatomical point of view we cannot recommend this position.^59,64 The “over the top position” is too deep and too superior in the notch and does not correspond to the anatomical PL bundle location. Rosenberg and Graf⁴⁵ recommend placing the PL bundle directly next to the AM bundle. One should keep in mind not every “double bundle technique” seems to replicate the anatomy of the ACL. Therefore, the term “anatomic ACL reconstruction” should only be used for techniques with the tunnel exactly at the origin and insertion of the AM and PL bundle.

For cases in which the ACL attachment on the femur cannot be identified we particularly recommend the coordinates given by Yasuda et al⁶⁰ and Zantop et al⁶⁴ for the PL bundle location. From the anatomical viewpoint it is essential to visualize the PL origin via a medial portal as recommended by Yasuda et al⁶⁰ or Zantop et al.⁶⁴

There is also some disagreement on how to drill the PL tunnel (transtibial or medial portal technique). Yasuda et al⁶⁰ recommend the transtibial drilling, whereas others recommend the medial portal technique.^8,13,55 From the anatomical point of view the origin of the PL bundle can be reached with both techniques. However, one has to keep in mind that with the transtibial technique the PL bundle origin should be considered when drilling the tibial tunnel. The medial portal provides more flexibility for choosing the PL bundle origin. It seems clinically relevant that the PL origin has a close topographic relation to the articular cartilage of the lateral femoral condyle. If the medial portal for drilling the PL tunnel is too high or the flexion angle is too low the drill has the potential to damage the subchondral bone plate or the articulate cartilage.

It is generally agreed ACL reconstruction using two tunnels at the femur and at the tibia is technically demanding.^8,13,55,60 To reproduce the anatomy, it is mandatory to place the tunnels exactly within the femoral origin and tibial insertion of the AM and PL bundle. One should always keep in mind if the first tunnel is placed in the center of the ACL attachment instead of in the center of the anteromedial or posterolateral bundle, there is no chance for proper placement of the second tunnel. At the tibial side the AM tunnel should be placed more anteriorly than a single tunnel. This means there is a higher chance for notch impingement. However, one should keep in mind there is a physiological impingement between the AM bundle and the anterior notch. To prevent a pathological notch impingement it is mandatory to place the femoral AM tunnel in the normal AM origin and not at the notch roof and to mimic the tension pattern of the normal ACL. For proper tunnel placement the use of fluoroscopy or the use of a navigation system could be of help. For both methods the radiological coordinates given in the present study are needed.

We believe the double bundle reconstruction of the ACL mimics the anatomy more closely than a single bundle reconstruction. However, this technique is associated with several pitfalls which can be avoided by having comprehensive knowledge of the anatomy of the AM and PL bundle. By understanding the anatomic details the double bundle technique should not lead to double trouble as suggested in the recent editorial by Harner and Poehling.²⁵