Scientific Background: Anatomy and Biomechanics

The anterior cruciate ligament (ACL) consists of dense connective tissue enveloped in a synovial membrane, which places the ligament in an intra-articular but extrasynovial position.^2,5 It attaches proximally on the posterior aspect of the lateral femoral condyle (LFC) and runs in an oblique course distally through the intercondylar notch to insert between the medial and the lateral tibial spines. Many authors have studied the ACL bundle anatomy and reached a general consensus that the ACL consists of two bundles: an anteromedial (AM) bundle and a posterolateral (PL) bundle (Fig. 10-1A). The AM bundle is slightly larger in diameter than the PL bundle. The bundles are named for their relative positions on their tibial insertion sites⁴ (see Fig. 10-1B). Recently, the two-bundle anatomy was also verified in a fetal study by Ferretti and coworkers.⁶

FIGURE 10-1 The anterior cruciate ligament (ACL) consists of two functional bundles, the anteromedial (AM) bundle and the posterolateral (PL) bundle. A, Arthroscopic view of the AM and PL bundle. B, Cadaveric dissection of tibial AM and PL insertion site.

The authors have studied the bony topography of the femoral attachment of ACL extensively. Using fetal specimens, cadavers, and in vivo arthroscopic observation, we have identified two osseous ridges that define the origins of the AM and PL bundles. The lateral intercondylar ridge runs proximal to distal through the entire ACL femoral attachment. With the knee in extension, no fibers of the ACL are attached anterior to this ridge. A second osseous ridge, the lateral bifurcate ridge, divides the femoral attachments of the AM and the PL bundles. It is important to note that when the knee is in full extension (anatomic position), the femoral origin of the AM bundle is located at the posterior and proximal portion of the lateral intercondylar wall, whereas the origin of the PL bundle is located slightly distally. The two bundles are parallel in extension. As the knee is flexed to 90°, which is the typical position during ACL reconstruction, the origins of the two bundles change from a vertical alignment to a horizontal alignment and the bundles cross.⁴ The lateral intercondylar ridge delineates the superior border of ACL femoral attachment (Fig. 10-2), whereas the lateral bifurcate ridge runs from superior to inferior and separates the femoral AM and PL attachments.

FIGURE 10-2 The lateral intercondylar ridge (arrows) marks the superior aspect of the femoral insertion sites of the AM and PL bundles when the knee is in 90° of flexion. The lateral bifurcate ridge (arrowheads) separates the insertions between the two bundles.

Biomechanically, the two bundles are not isometric throughout the range of knee motion. Generally, the AM bundle maintains at a constant level of tension throughout the range of motion, with some increase when the knee is flexed that reaches a maximum at 60°.⁷ The tension of the PL bundle is more variable because it tightens in knee extension and slackens in flexion past 30°. Thus, the AM and PL bundles have varying contributions to knee stability at different flexion angles. The AM bundle limits anteroposterior (AP) translation throughout knee motion, whereas the PL bundle plays an important role in limiting not only anterior tibial translation but also rotation.^3,7,10

Biomechanical studies have emphasized the importance of both bundles in knee stability. Yagi and associates¹³ showed that double- bundle (DB) ACL reconstruction better restores knee biomechanics than single-bundle (SB) ACL reconstruction. The addition of a PL bundle produces in situ forces within each bundle that closely match the in situ forces found in a native ACL ligament. Tashman and colleagues¹¹ studied the in vivo kinematics after ACL reconstruction with the use of high-speed stereoradiography. This study demonstrated that SB ACL reconstruction sufficiently restored AP tibial stability. An unexplained increase of 3° to 4° in adduction and external tibial rotation was reported. Zantop and coworkers¹⁵ showed that isolated transection of the AM bundle increased anterior tibial translation at 60° and 90° of knee flexion significantly, whereas isolated transection of the PL bundle significantly increased anterior tibial translation at 30° of flexion. In addition, PL bundle transection led to significantly increased rotation at 0° and 30° in response to a combined rotatory load when compared with the intact knee and the AM bundle–deficient knees. This study supports the concept that SB (AM bundle) reconstruction cannot restore native knee stability, particularly rotatory stability.

Primary Knees

DB ACL reconstruction is indicated for acute or chronic ACL ruptures that cause symptomatic instability. The criteria for DB ACL reconstruction are (1) a history of instability, shifting, or giving-way with activities; (2) positive Lachman and pivot shift tests; (3) increased KT-1000 measurement compared with the contralateral knee; and (4) patient desires to undergo ACL reconstruction. Whereas some surgeons reserve the DB procedure for patients with global ligamentous laxity or very large pivot shifts, the authors currently advocate the procedure for most patients with clinical instability from an ACL tear.

Revision Knees

DB ACL reconstruction as a revision procedure may be indicated when an SB ACL reconstruction has failed owing to graft rupture or when an SB graft is intact but does not provide adequate stability. In the latter case, depending on the position of the original tunnels, the surgeon may either augment the intact SB graft with an additional graft or implant two new grafts. In the authors’ practice, patients frequently present with an intact SB graft but complain of clinical symptoms of instability and demonstrate a positive pivot shift on examination. In these cases, there is often a mismatch in the femoral and tibial tunnel placements. For example, the tibial tunnel occupies the position of the PL bundle and the femoral tunnel is located in a high position in the notch above the AM bundle attachment (Fig. 10-3). This pattern of mismatch is often a result of the transtibial technique for femoral tunnel preparation, in which the femoral tunnel position is dictated by the tibial tunnel position. The graft in this case may provide stability in the AP plane but does not provide adequate rotational stability and results in symptomatic instability.

FIGURE 10-3 This sagittal magnetic resonance imaging (MRI) scan shows an example of tunnel mismatch. The tibial tunnel occupies the position of the PL bundle. The femoral tunnel is closer to the attachment of the AM bundle.

A nonanatomic tunnel placement, or tunnel mismatch, may lead to graft laxity and ultimate failure. In the laboratory, increased osteoclastic activity has been observed at the graft-bone interface in such mismatched tunnels (Fig. 10-4). When a graft is placed at a nonanatomic position relative to the native ACL, it experiences large tensile forces secondary to this abnormal position. These abnormal forces compromise biologic healing and ultimately may contribute to graft laxity and failure.

FIGURE 10-4 Osteoclasts are shown at the graft–bone tunnel interface. We have observed increased osteoclastic activity at the graft-bone interface in cases of tunnel mismatch.

Primary Knees

There are certain contraindications to DB ACL reconstruction. Patients with ACL rupture who do not have symptomatic instability and do not participate in cutting and pivoting sports may choose not to undergo ACL surgery, because they can function well without an intact ligament. Patients with a limitation of knee motion and flexion contractures must first be rehabilitated appropriately with aggressive physical therapy before consideration of DB ACL reconstruction. Patients with active infection are not candidates for reconstructive surgery. Skeletally immature patients are not candidates for DB ACL reconstruction owing to the risk of growth arrest from transphyseal tunnel drilling.

Revision Knees

Contraindications to DB ACL reconstruction in the revision setting include active infection and skeletal immaturity. Excessive tunnel widening and bone loss from prior ACL reconstruction surgery necessitates staging of the revision procedure. In these cases, bone-grafting of the femoral and tibial tunnels is first performed. After the bone graft has healed, revision DB ACL surgery is accomplished.

Primary Knees

The first portion of the clinical evaluation consists of obtaining a thorough history from the patient. The timing and description of the injury mechanism should be reviewed carefully because this information may raise suspicion of injury to particular structures depending on the direction of force to the knee. The patient should be asked if he or she felt or heard a pop at the time of the injury. In addition, the development of an effusion and difficulty with weight-bearing should be noted. Episodes of recurrent shifting or instability are important to document, including the situations in which they occurred. Any previous treatment of the injury such as physical therapy or bracing should be documented. The patient’s preinjury activity level should be determined to assess if he or she engages in activities, such as soccer or basketball, that require cutting and pivoting motions. Any previous history of knee injury or knee surgery should be explored.

The second portion of the clinical evaluation is a comprehensive physical examination. Weight-bearing alignment is evaluated for any valgus or varus deformity. Before the patient sits or lies on the examining table, it is important to assess her or his gait for the presence of a limp and note whether any assistive devices are required for ambulation. With the patient supine, the presence and size of an effusion is noted. Quadriceps function and atrophy are assessed, including the patient’s ability to perform a straight leg raise without an extensor lag. Joint range of motion is measured and compared with that of the contralateral knee. The patellofemoral joint is examined for apprehension, crepitation, and facet tenderness. The ligament examination includes the Lachman, pivot shift, anterior drawer, posterior drawer, and valgus and varus laxity tests. The ligament examination is performed bilaterally for comparison. Joint line tenderness and the McMurray flexion test are done to detect meniscal pathology. KT-1000 testing of both knees is performed and the difference in millimeters recorded.

When an ACL injury is suspected, diagnostic studies are ordered. Radiographs include a weight-bearing posteroanterior flexion view, a lateral view of the injured knee, and a Merchant view of both knees. In patients with abnormal alignment or previous surgery or who are older, long-leg cassette films are obtained to quantify malalignment. Radiographs are scrutinized for joint space narrowing, osteochondral lesions, fractures, and alignment abnormalities.

The other diagnostic study of importance is magnetic resonance imaging (MRI). At the authors’ center, a special protocol to study the ACL consists of a series of images in the plane of the two bundles of the ACL (Fig. 10-5A) in addition to the usual coronal, axial, and sagittal images. This provides improved visualization of the individual AM and PL bundles, allowing identification of individual bundle ruptures (see Fig. 10-5B) and the location of the rupture as either off the femur or the tibia or in mid-substance. In addition, assessment is performed of the condition of the other knee ligaments, menisci, cartilage, and bony structures.

FIGURE 10-5 A, An example of the special MRI scan in the oblique plane of the ACL clearly demonstrates the presence of intact AM and PL bundles. B, MRI scans in the plane of the ACL allow identification of individual bundle rupture. In this case, the PL bundle is intact and the AM bundle is torn.

Revision Knees

Clinical evaluation of a patient with a failed previous ACL reconstruction is identical to that of a patient with a primary ACL rupture, including a complete history and comprehensive physical examination. It is important to ascertain the mechanism of graft failure as well as the timing in relation to the primary procedure. Some patients will recall a specific episode in which the knee was reinjured. Others will not recall a reinjury but will state that the knee never felt stable even after the initial ACL reconstruction. Another important component of the history includes the operative report from the primary surgeon to identify the type of graft used, fixation methods, and tunnel widths.

A complete physical examination is performed, as previously described. The surgeon must pay specific attention to previous incisions and sites of graft harvest in order to plan the revision appropriately. In some cases, the Lachman and KT-1000 measurements are not as impressive as the pivot shift, indicating the presence of a nonanatomic graft that provides AP stability but not rotational stability.

Radiographs are required to determine fixation methods used in the primary procedure (Fig. 10-6A). Joint space narrowing, osteochondral lesions, and tunnel width are identified. MRI is performed to determine whether the original graft is intact or ruptured, assess tunnel location and width (see Fig. 10-6B and C), and identify additional pathology involving the cartilage or menisci.

FIGURE 10-6 Anteroposterior (AP) radiograph (A), sagittal (B) and coronal images (C) from a patient with a failed single-bundle ACL reconstruction. A and C, Vertical femoral tunnel placement is noted. B, The graft is shown to insert in the area of PL bundle insertion on the tibia. On the femoral side, the graft inserts anteriorly and superiorly to the AM bundle femoral insertion. A revision ACL reconstruction using the double-bundle technique was performed in a single-stage procedure.

Primary Knees

Once the ACL rupture has been diagnosed and DB ACL reconstruction selected, preoperative planning is conducted that includes a careful review of any associated injuries, such as meniscal tears, and plan for treatment of these associated injuries. Graft options are reviewed and discussed with the patient. The authors routinely use two tibialis anterior allografts for this procedure. The operation and postoperative rehabilitation are reviewed with the patient prior to surgery.

Revision Knees

Preoperative planning takes on special importance in the case of revision DB ACL surgery. The surgeon must review the patient’s prior operative report and arthroscopic images, if available. From the physical examination and imaging, the integrity of the graft and the location of the femoral and tibial tunnel positions in relation to the anatomic attachments of the AM and PL bundles are determined. Tunnel mismatch is assessed. The overall orientation of the graft (e.g., vertical) is determined from the coronal and sagittal images of the MRI. If tunnel widening is severe, the surgeon should stage the revision ACL surgery by first bone-grafting the old tunnels.

If the graft is intact and there is no tunnel mismatch, an appropriate plan may entail augmentation of the old graft with a new graft. If the graft is ruptured, or the tunnels are nonanatomic, the graft is removed and new femoral and tibial tunnels drilled. The surgeon must be aware of the limited space available for creation of new tunnels and try to determine preoperatively whether the previous tunnel will interfere with anatomic tunnel placement. In some cases, an over-the-top location for the femoral AM graft is required owing to limited space on the femoral side. However, use of the transtibial technique in the primary procedure frequently results in an excessively high femoral tunnel. This tunnel is usually well superior to the anatomic AM bundle attachment site and does not preclude placement of new anatomic AM and PL femoral tunnels during revision DB ACL surgery.

Primary Knees

While the patient’s limb is being positioned, prepared, and draped, an assistant begins graft preparation. All allografts are obtained from a tissue bank certified by the American Association of Tissue Banks and inspected by the U.S. Food and Drug Administration (Lifenet, Virginia Beach, VA). The tibialis anterior grafts are thawed and rinsed on the back table in antibiotic solution. The allografts are frequently 24 to 30 cm in length and, when doubled over, provide 12- to 15-cm double-stranded grafts. The tendon allografts are trimmed of excess or loose tissue. When folded over, the AM graft is approximately 8 mm in diameter and the PL graft is approximately 7 mm (Fig. 10-8). Smaller-diameter grafts may be chosen in patients with small knees. The ends of the grafts are sutured using a baseball stitch with No. 2 Ti-Cron sutures for a length of 2.5 cm. An EndoButton CL (Smith & Nephew, Andover, MA) is applied to each graft, resulting in a double-stranded graft. The authors most commonly use a 15-mm-long loop. In cases in which patients refuse the use of allograft tissue, a hamstring autograft may be used. The hamstrings are prepared in the same fashion as the tibialis anterior allografts. The semitendinosus tendon is used for the AM bundle and the gracilis tendon is used for the PL bundle. Typically, the diameter of each hamstring autograft is smaller than that of the allograft tissue. Usually, the semitendinosus tendon graft measures 7 to 8 mm and the gracilis tendon graft, 5 to 6 mm. The authors prefer allograft tissue owing to consistency of size to restore the entire footprint of the AM and PL bundles of the ACL.

FIGURE 10-8 For grafts, we commonly use two tibialis anterior allografts. Doubled over, the AM bundle has a diameter of 8 mm, and the PL bundle, 7 mm.

Arthroscopy is performed for diagnosis and treatment of associated injuries. Placing the portals in the correct location is critical for visualization and technical ease during the procedure. First, the anterolateral (AL) portal is established off the lateral edge of the patellar tendon at the level of the inferior pole of the patella. This portal is placed slightly higher than a traditional AL portal to obtain a broader view with the arthroscope. This portal is created with a No. 11 blade with the knee at 45° of flexion to avoid injuring the patella. An AM portal is required, along with an accessory anteromedial portal (AAM) (Fig. 10-9). The AM portal is placed along the medial border of the patellar tendon, inferior to the level of the AL portal. To establish the AAM portal, the arthroscope is placed in the AM portal, and an 18-gauge spinal needle is inserted medial and distal to this portal just above the medial meniscus. The trajectory of the needle aims to reach the center of the femoral footprint of the PL bundle. Care must be taken to ensure that the spinal needle does not pierce the medial meniscus and avoids contact with the medial femoral condyle because reamers must pass through this portal without damaging the medial femoral condyle. Once the needle is placed in the correct position, the AAM portal is made with a No. 11 blade. With the camera in the AL portal, a portion of the fat pad is débrided to allow visualization of the ACL insertions. Associated injuries are addressed before the ACL reconstruction.

FIGURE 10-9 We use three portals for double-bundle ACL reconstruction. In addition to an anteromedial (AM) and anterolateral (AL) portal, an accessory anteromedial (AAM) portal is established, slightly inferior and medial to the AM portal. This portal affords the appropriate trajectory to reach the femoral footprint of the PL bundle as well as the AM bundle during femoral tunnel preparation.

The rupture pattern of the ACL is evaluated by carefully probing the ligament. A meticulous, gentle dissection of the ACL bundles is performed using a thermal device (Arthrocare Corporation, Sunnyvale, CA), taking great care to preserve the soft tissue remnants at the tibial and femoral attachments and any intact bundle fibers. The determination is made whether one or both bundles are torn and whether the tear involves the femoral or tibial attachment. If one of the bundles is intact, it is preserved and augmented by reconstructing the ruptured bundle only. Preserving the soft tissue attachments allows anatomic identification of the proper locations for tunnel placement. Placing the arthroscope in the AM portal allows improved visualization of the femoral footprints. Furthermore, placement of the arthroscope in the AM portal makes notchplasty unnecessary in almost every case because of the excellent view obtained through this portal (Fig. 10-10).

FIGURE 10-10 Placing the arthroscope in the AM portal allows excellent visualization of the femoral insertions of the AM and PL bundles. Viewing through the anterolateral portal does not allow this view because the lateral condyle blocks visualization unless a notchplasty is performed.

In chronic cases, the ACL rupture pattern and insertion sites may not be as visible compared with acute cases. A strong understanding of ACL anatomy is essential for accurate identification of the footprints of the bundles. The identification of two ridges, the lateral intercondylar ridge and the lateral bifurcate ridge previously described, assists to guide femoral tunnel placement (see Fig. 10-2). With the knee at 90° of flexion, the lateral intercondylar ridge (also known as the resident’s ridge) marks the superior border of the AM and PL bundles. No ACL tissue attaches above this ridge. The lateral bifurcate ridge is a bony prominence that divides the AM and PL insertions. When the knee is flexed 90°, the AM bundle femoral insertion is located posterior to the lateral bifurcate ridge and the PL bundle femoral footprint is located anterior to the lateral bifurcate ridge. Once identified, the femoral footprints of both the AM and the PL bundles are marked with the thermal device. A curved Steadman awl is used to create a small hole in the center of the femoral AM and PL footprints for subsequent guidewire placement. The tibial footprints are also marked by the thermal device. Both the tibial and the femoral footprints are measured for length and width to determine tunnel diameter and graft size (Fig. 10-11).

FIGURE 10-11 The tibial footprints are measured for length (A) and width (B). The femoral footprints are also measured for length (C) and width (D).

The PL femoral tunnel is prepared first. With the arthroscope in the AM portal, a 3.2-mm guidewire is inserted through the AAM portal and placed in the center of the femoral footprint of the PL bundle previously defined by the Steadman awl. The PL femoral tunnel pin is placed 5 to 7 mm from the anterior LFC cartilage and 3 mm from the inferior LFC cartilage. During PL femoral tunnel preparation, the knee is flexed to at least 110° to protect the peroneal nerve and to ensure adequate EndoButton tunnel length. The same flexion angle must be maintained throughout tunnel preparation and resumed for PL graft passage. Once the guidewire is placed into position, a 6-mm cannulated acorn drill is placed over the guidewire and used to ream the tunnel to a depth of 25 to 30 mm. The far cortex is breached with a 4.5-mm EndoButton drill, and the depth gauge is used to measure the distance to the far cortex. The depth of reaming to allow the Endobutton drill to flip is calculated and the tunnel is reamed by hand to that depth with a 7-mm reamer.

Next, the AM and PL tibial tunnels are prepared. A 3- to 4-cm skin incision is made over the AM surface of the tibia at the level of the tibial tubercle. The PL tibial tunnel is drilled first. With the arthroscope in the AL portal, the elbow ACL tibial drill guide is set at 45° and the tip of the drill guide is placed on the tibial footprint of the PL bundle via the AAM portal. The tibial PL bundle footprint is adjacent to the posterior root of the lateral meniscus and posterolateral to the AM bundle of the ACL (Fig. 10-12). On the tibial cortex, the tibial drill for the PL tunnel starts just anterior to the superficial medial collateral ligament fibers. Once the tibial drill guide is set, a 3.2-mm guidewire is passed into the base of the PL tibial footprint. The AM tibial tunnel is similarly drilled with the elbow ACL tibial drill guide set at 45° inserted through the AM portal. On the tibial cortex, the starting point for the AM bundle is midway between the PL starting point and the tibial tuberosity. There should be an adequate distance between these two pins to ensure sufficient bony bridge between the tunnels (Fig. 10-13). The AM and PL tibial tunnels are then overdrilled using a cannulated drill. The PL tunnel is usually accomplished with a 6-mm drill and the AM with a 7-mm drill. A curette is placed over the tip of the guidewire during reaming to protect the femoral articular cartilage. A dilator is used to widen the tunnels to the specific diameters, which are most commonly 8 mm for the AM tunnel and 7 mm for the PL tunnel.

FIGURE 10-12 The PL bundle inserts on the tibia adjacent to the posterior horn attachment of the lateral meniscus.

FIGURE 10-13 Guidewires for the AM and PL tibial tunnels are placed, centered in the insertions of the AM and PL bundles.

The femoral AM tunnel is the last tunnel to be drilled. The arthroscope is placed in the AM portal for this part of the procedure. First, a transtibial approach is attempted with a guidewire placed through the AM tibial tunnel (Fig. 10-14A). However, in some cases, the correct position for the AM femoral tunnel cannot be reached with this technique. Often, this places the guidewire too high and too far anterior on the medial wall of the lateral femoral condyle, outside of the anatomic footprint. In these cases, a transtibial technique through the PL tibial tunnel is attempted (see Fig. 10-14B). In some cases, the trajectory of the guidewire is still too high. When the center of the AM femoral footprint cannot be reached through either the AM or the PL tibial tunnel, the guidewire is placed through the AAM portal (see Fig. 10-14C). Usually, the knee must be flexed greater than 90° to position the pin appropriately in the center of the AM bundle attachment.

FIGURE 10-14 The guidewire can be inserted in one of three ways to reach the AM femoral attachment site for AM femoral tunnel preparation: transtibially through the AM tibial tunnel (A), transtibially through the PL tibial tunnel (B), or through the accessory anteromedial (AAM) portal (C).

With the tip of the guidewire in place at the center of the AM femoral footprint, a cannulated acorn drill is inserted over the guidewire, drilling to a depth of 20 to 30 mm. A smaller depth is used when drilling through the AAM portal than transtibially owing to the shorter distance to the femoral cortex. The far cortex of the AM femoral tunnel is breached with a 4.5-mm EndoButton drill, and the depth gauge is used to measure the distance to the far cortex. The appropriate depth for the EndoButton to flip is calculated (the tunnel length minus 7 mm if a 15-mm loop is used), and the tunnel is reamed by hand to this depth with an 8-mm reamer. An assistant marks both the AM and the PL grafts with a marking pen at the appropriate distance from the EndoButton for the tunnel length. The final appearance of the tunnels after completion of preparation is shown in Figure 10-15.

FIGURE 10-15 The final appearance of the tunnels prior to graft passage, with dilators protruding from the AM and PL tibial tunnels.

The next step is graft passage, beginning with the PL graft. A Beath pin with a very long looped suture in the eyelet is passed through the AAM portal and out through the PL femoral tunnel. The looped suture is visualized intra-articularly and retrieved with an arthroscopic suture grasper through the PL tibial tunnel. This process is repeated for the AM graft, passing the Beath pin through whichever tunnel or portal was used to drill the AM femoral tunnel, and retrieving it through the AM tibial tunnel. The sutures from the Beath pin cross over each other with the knee in flexion (Fig. 10-16). Both Beath pins are passed before inserting the PL graft into the joint to protect it from injury. Subsequently, the PL graft is passed through the PL tibial tunnel into the PL femoral tunnel. Once the graft markings are seen, the EndoButton is flipped. During PL graft passage, the knee must be hyperflexed, just as it was during PL femoral tunnel preparation.

FIGURE 10-16 When both Beath pins have been passed into the AM and PL tunnels, the suture loops from the Beath pins cross over each other at 90° of knee flexion.

After passing the PL graft, the AM graft is passed retrograde and the EndoButton is flipped onto the lateral femoral cortex (Fig. 10-17). Preconditioning of the grafts is performed by flexing and extending the knee from 0° to 120°, approximately 20 to 30 times. Fixation on the tibial side is performed with bioabsorbable screws that are the same size in diameter as the grafts. We don’t actually measure the tension, but manually tension the graft and then confirm adequate tension by checking Lachman test and graft tension intraoperatively after fixation. The PL graft is fixed with the knee in extension, and the AM graft is fixed at 60° of flexion. After tibial fixation, a final arthroscopic inspection is performed to confirm the correct position and tension of the grafts. The surgeon verifies that there is no roof impingement or posterior cruciate ligament (PCL) impingement. Interestingly, these phenomena have not been observed in the setting of an anatomic DB ACL reconstruction.

FIGURE 10-17 The final appearance after AM and PL graft passage shows a crossing pattern when in flexion. Also, the PL bundle is mostly covered by the AM bundle.

Subcutaneous tissue and skin are closed in layers. A dry sterile dressing is applied. A Cryo-Cuff (Aircast, Vista, CA) is used for cold therapy and compression. A hinged knee brace is applied and locked in extension for ambulation.

Rehabilitation for anatomic DB ACL reconstruction is the same as that used for SB ACL reconstruction. Weight-bearing with crutches is allowed immediately after surgery, progressing from partial to full. During the first 1 to 3 months postoperatively, restoration of normal range of knee motion and recovery of muscle strength are emphasized. Closed-chain and functional exercises are incorporated, as well as balance and neuromuscular training. Plyometric and sports-specific exercises are begun 3 months postoperatively. The patient is allowed to gradually return to sports activities at approximately 6 months postoperatively. The individual should have no pain or swelling, 80% to 90% quadriceps strength, a normal single-leg hop test, and adequate proprioception and neuromuscular control before being fully released back to sports.

Revision Knees

The AL, AM, and AAM portals are established as previously described. A careful arthroscopic examination is performed to assess the position and integrity of the graft. The graft may be intact but quite lax and stretched out, rendering it nonfunctional. If the graft is intact and functional, the location of the femoral and tibial tunnels must be judged. If the intact, functional graft is in an acceptable position for an AM or PL bundle, an augmentation procedure is performed in which the absent bundle is reconstructed, leaving the intact graft alone. Commonly, the tunnels for the intact graft are mismatched. Most often, the tibial tunnel is located in the area of the PL bundle attachment, and the femoral tunnel is placed anterior and superior to the AM bundle attachment (Fig. 10-18). In this case, the graft must be resected and new femoral tunnels established. Depending on the size of the original tibial tunnel, either one large tunnel for both AM and PL bundles or two separate tunnels may be used. If tunnel widening is excessive, the surgeon should consider bone-grafting the tunnel and staging the revision DB ACL reconstruction.

FIGURE 10-18 A, A three-dimensional computed tomographic rendering of the distal femur in a failed single-bundle ACL reconstruction. The rendering demonstrates the nonanatomic position of the previous femoral tunnel, well superior to the native AM bundle attachment. B, A similar rendering of the proximal tibia in the same case of failed single-bundle ACL reconstruction. The tibial tunnel occupies the PL tibial insertion site, thus creating femoral and tibial tunnel mismatch.

Once the surgeon has determined that the graft is either incompetent or nonanatomic, the fixation devices on the tibial and femoral side are addressed. Staples and screws may be removed from the tibial side through the old incision to facilitate tibial tunnel preparation. On the femoral side, if the tunnel is nonanatomic and does not interfere anatomically with the creation of two new tunnels, the fixation device is left alone.

If the graft is incompetent or nonanatomic, it is resected with a thermal device. The thermal device is used to perform a gentle dissection of any remaining soft tissue to identify locations for the new femoral tunnels. In primary DB ACL surgery, the lateral intercondylar ridge and lateral bifurcate ridge are helpful bony landmarks to guide AM and PL femoral tunnel placement. In revision DB ACL surgery, these landmarks are often not easily identified. Frequently, a notchplasty was performed at the time of the primary ACL reconstruction, removing these landmarks and distorting the anatomy. In these cases, the surgeon must rely on his or her knowledge of ACL insertional anatomy to guide tunnel placement in the revision procedure.

As described for primary DB ACL reconstruction, a Steadman awl is used to mark the center of the eventual AM and PL tunnels on the lateral femoral wall. Placing the arthroscope in the AM portal facilitates visualization of the wall and allows accurate tunnel markings. Attention is turned first to preparation of the PL femoral tunnel, which is done in the same manner as previously described.

Next, attention is turned to tibial tunnel preparation. If the original tibial tunnel is small enough in relation to the size of the knee and placed near the insertion of either the AM or the PL bundle, it is theoretically possible to drill a separate new tibial tunnel. However, in many revision cases, the original tibial tunnel is enlarged, occupying the area of both AM and PL bundle attachments. Therefore, two separate tibial tunnels cannot be used because both AM and PL grafts will traverse a single tibial tunnel. The old tibial tunnel is reamed and all remaining soft tissue removed from the walls of the tunnel. The tunnel is then expanded to accommodate both the AM and the PL grafts, attempting to create a figure-of-eight–type aperture that allows an anterior recess for the AM graft and a posterior recess for the PL graft. If the position of the original tibial tunnel is completely unsatisfactory, the surgeon should strongly consider bone-grafting the tunnel and staging the revision DB ACL procedure.

The AM femoral tunnel is prepared. In many cases, the original femoral tunnel is superior to the AM femoral insertion and, therefore, does not interfere with drilling a new, anatomic AM femoral tunnel. In this case, the AM femoral tunnel is prepared as described for a primary DB ACL procedure. Upon completion of tunnel preparation, the grafts are passed as described for a primary DB ACL reconstruction. If a single tibial tunnel is used, the surgeon must use a probe during graft passage to maintain the relative anterior and posterior positions of the AM and PL grafts.

In some cases, the original tunnel encroaches upon the AM femoral insertion and there may not be enough room to drill a new AM femoral tunnel. An over-the-top procedure will circumvent this problem and allow near-anatomic graft placement. For the over-the-top procedure, the back wall of the lateral femoral condyle is rasped to create a healing surface for the graft. Next, an incision is made along the lateral aspect of the thigh near the lateral femoral condyle. Dissection is carried down to the level of the vastus lateralis, which is elevated superiorly with a retractor to expose the over-the-top position on the posterior aspect of the lateral femoral condyle. Next, a looped 18-gauge wire is passed on a curved clamp from the intercondylar notch around to the over-the-top position. This wire is retrieved through the lateral incision. Another looped wire is placed through the original wire, and the original wire is retrieved through the notch, thereby pulling the looped wire anterograde into the notch. The loop is retrieved with an arthroscopic grasper through the tibial tunnel.

Once the wire loop is positioned in the tibial tunnel, the Beath pin for the PL tunnel is passed as previously described. The PL graft is passed first, and the EndoButton deployed on the outer femoral cortex. Next, the sutures on the AM graft are passed through the wire loop, and the wire is pulled out of the lateral incision, bringing the AM graft into the tibial tunnel and into the over-the- top position on the femur. The graft is secured with staples, a screw, and a spiked washer or tied around a post on the distal femur.

If two separate tibial tunnels are used, the grafts are tensioned separately as described for a primary DB ACL procedure. In the revision setting, supplementary fixation with staples or a post is used in addition to a biointerference screw. If a single tibial tunnel is used for both grafts, tensioning of the grafts becomes more complex. Both bundles are manually tensioned at 20° of flexion and fixed with an interference screw in the single tibial tunnel. We don’t actually measure the tension of the graft. But the graft is manually tensioned and then adequate tension is confirmed by checking Lachman test and graft tension intraoperatively after fixation. Alternatively, the surgeon can tension and fix each graft separately, using either staples or sutures tied around a post. An interference screw may be added to the single tibial tunnel after individual bundle fixation if supplementary fixation is required.

Graft preparation for a revision DB ACL reconstruction is the same as that described for a primary DB ACL reconstruction. The authors commonly use two tibialis anterior allografts and EndoButtons for fixation on the femoral side. If an over-the-top procedure is performed, a long graft is required. For the over- the-top procedure, a 40- to 50-mm Endoloop may be placed on the doubled-over tibialis anterior allograft and the Endoloop fixed to the distal lateral femur around a screw and washer.

Primary Knees

From November 2003 to October 2007, one of the authors (F.F.) performed 489 anatomic DB ACL reconstructions. In a total of 393 primary cases, 374 patients (95%) underwent primary DB reconstruction and 19 patients (5%) underwent primary SB augmentation. Primary SB augmentation was performed when one of the bundles was found to be intact. In these cases, the intact bundle was augmented by reconstruction of the ruptured bundle.

A prospective study was done to evaluate the outcomes of anatomic DB ACL reconstruction. Lachman and pivot shift tests, KT-2000 testing, range of knee motion, and overall International Knee Documentations Committee (IKDC) rating were accomplished. Compared with primary SB ACL reconstruction,^6a knees that had primary DB ACL reconstruction demonstrated superior range of knee motion at 1, 4, and 12 weeks postoperatively. All patients who undergo DB ACL reconstruction will be followed prospectively for long-term results to document range of knee motion, ligamentous laxity, functional strength, and activity and sports participation.

At the time of writing, the initial 100 consecutive patients who underwent primary DB ACL reconstruction had been followed an average of 2.1 ± 0.5 years postoperatively. None of these patients had undergone concomitant ligament or cartilage restorative procedures. Fifteen had a concomitant medial meniscus repair, 14 had a partial lateral meniscectomy, 9 had a partial medial meniscectomy, and 1 had a lateral meniscus repair.

Evaluation revealed a mean side-to-side difference in range of motion of 2° ± 3° for extension and 2° ± 5° for flexion. One patient lacked 15° of extension and another lacked 13° of extension compared with the opposite knee. In both, the contralateral knee had a large amount of hyperextension (–10° and –13°) and a manipulation under anesthesia was required to restore hyperextension.

The Lachman test was normal in 65% and nearly normal in 33%. The pivot shift test was normal in 94% and nearly normal in 6%. KT-2000 testing had been done in 87 patients at the time of writing. The mean side-to-side difference was 1.0 ± 2.3 mm. Fifty-one patients had less than 3 mm of increased anterior translation, 32 patients had 3 to 5 mm of increased translation, and 4 patients had greater than 5 mm of increased translation. There were 8 graft failures, 7 of which underwent subsequent revision surgery.

No patient had complaints of pain, swelling, or instability during activities of daily living, and 73% to 78% had no symptoms during very strenuous or strenuous sports activities. The IKDC Subjective Knee Form and Knee Osteoarthritis Outcome Score Activities of Daily Living and Sports Activity scores were 85.0, 91.8, and 87.0, respectively. These scores were similar in comparison with patients undergoing SB ACL reconstruction previously reported. Fifty-one percent of the patients described their current activity level as normal and 35% as nearly normal. From these initial results, the conclusion was reached that anatomic DB ACL reconstruction results in good restoration of joint stability and patient-reported outcomes when evaluated 2 years after surgery.

Revision Knees

In a total of 100 revision cases, 73 patients underwent DB ACL revision and 27 patients had secondary SB augmentation. For secondary augmentation cases,^4a 23 patients had previous SB reconstruction and complained of instability. In spite of good range of motion and minimal side-to-side differences on KT-2000 testing, they were unstable during pivot shift testing. During surgery, the original AM graft was found to be functionally intact, and therefore, only PL bundle reconstruction was performed. The other 4 secondary augmentation patients had previous DB ACL reconstruction. The AM graft was shown to be stretched, and the PL graft remained functional. Therefore, only the AM graft was reconstructed; the PL graft was left intact.

Clinical and surgical data were prospectively collected on 16 patients with a failed SB ACL reconstruction that was revised to a DB reconstruction with a minimum 1 year follow-up.^4a All patients were clinically unstable preoperatively. The AM graft was placed either in the original femoral tunnel, in a new femoral tunnel, or in an over-the-top position on the femur. At latest follow-up of an average of 1.1 years, 14 patients had a grade I Lachman and all 16 patients had a grade I pivot shift test.