Background

Graft selection for anterior cruciate ligament (ACL) reconstruction has been debated in the literature.¹ Historically, the debate has been regarding the use of soft tissue grafts secondary to increased laxity as compared with patellar tendon autografts.² The point of concern occurred at the site of graft fixation with soft tissue grafts, and the issues were “suspensory” fixation and creep with cyclical loading. Improved fixation techniques have largely eliminated the historical differences in laxity comparing quadrupled hamstring tendon with bone–patellar tendon–bone (BPTB) autografts.³ It has been well documented in recent literature that the two graft types have similar success rates.^4–6 Some potential advantages of soft tissue grafts include decreased kneeling pain, decreased patellar tendonitis, lower risk of postoperative patella fracture, and improved cosmesis with a smaller incision for harvest.¹

The patellar tendon graft, with bone plugs on each end, has classically been fixed with a metal interference screw in the tibia and femur with near-aperture fixation. This fixation was sufficiently rigid to prevent significant graft slippage or creep. The patellar tendon graft became the gold standard by which all other grafts and graft fixation systems were compared.¹ Despite the increased laxity with hamstring tendon reconstructions, this technique gained popularity as patients complained of significant anterior knee and kneeling pain with patellar tendon autografts. The potential for patellar fracture also exists. Hamstring tendon autografts have less morbidity, although they do demonstrate some permanent hamstring weakness at high flexion angles.⁷ The clinical significance of this weakness remains unclear.

A quadrupled hamstring tendon graft has two significant biomechanical advantages. First, quadrupled hamstring grafts have a larger cross-sectional area when compared with patellar tendon grafts.⁷ The larger cross-sectional area results in more collagen and thus a stronger graft. The ultimate tensile strength of a quadrupled hamstring autograft is 4140N with a stiffness of 807 N/mm. This is compared to the patellar graft at 2977N and 455 N/mm, respectively. The native ACL has been tested at 2160N and 242 N/mm, respectively.¹ Second, hamstring tendon grafts require smaller bone tunnels than do grafts with bone plugs. These grafts heal circumferentially to the tunnel wall if the interference screw is eliminated. These advantages fueled the search for better soft tissue graft fixation devices.

The historical poor performance of soft tissue grafts cannot be compared with today’s graft fixation. Initial studies compared doubled hamstring tendons to patellar tendons. A quadrupled soft tissue graft is the optimal choice when hamstring tendons are used. Studies of soft tissue graft fixation preceded the use of transfixion devices. Kousa et al⁸ tested the pullout strength of the soft tissue fixation devices that are currently most used for hamstring grafts. On the femoral side, the Bone Mulch Screw (Arthotek, Warsaw, IN) (1112N), Endobutton-CL (Smith & Nephew, Endoscopy Division, Andover, MA) (1086N), and Rigidfix (Depuy Mitek, Westwood, MA) (868N) systems had the best fixation for soft tissue grafts.⁸ These were significantly stronger than patellar tendons fixed with an endoscopic interference screw (588N). On the tibial side, the fixation is much more secure. The IntraFix (Depuy Mitek) (1332N) and WasherLoc (Arthrotek) (975N) both had higher pullout strengths than a patellar tendon bone plug fixed with a metal interference screw (758N, 9- × 30-mm screw). Interference screws for soft tissue grafts are weaker, with pullout strengths from 201N to 665N depending on the system.⁸

The improved fixation systems for hamstring tendons, combined with strength and stiffness superior to the native ACL, make this an attractive graft option. One must challenge any concept of a BPTB autograft as the gold standard. Patient factors and desires and the surgeon’s preference should dictate which graft is chosen. Surgeons completing ACL reconstructions should be well versed in both techniques. Allograft reconstruction offers an additional soft tissue option. For those surgeons or patients desiring an allograft source, the tibialis anterior tendon has an ultimate tensile strength of 4122N and is an attractive alternative.⁹ We also use the technique described here when an allograft tibialis is selected.

Surgical Technique

Patients undergoing ACL reconstruction are treated in the outpatient surgical suite. General anesthesia is combined with the intraarticular injection of a local anesthetic containing epinephrine to obviate the need for a tourniquet. The leg holder is positioned in as proximal a position on the thigh as possible so as to allow more than 100 degrees of flexion during graft tunnel preparation and graft passage if necessary. A well-padded tourniquet is placed on the upper thigh. It is not routinely inflated during the procedure in order to minimize quadriceps inhibition postoperatively. An examination under anesthesia is performed prior to final positioning.

The knee is initially prepped with povidone-iodine (Betadine) and injected with 60 mL of 1% lidocaine with epinephrine (1:100,000). The injection is performed laterally adjacent to the superior lateral portion of the patella. During the injection, a fluid wave should be visualized medially, which ensures that a fat pad injection has not occurred. The hemostatic effect would largely be lost with a fat pad injection. The knee is then prepped and draped with standard arthroscopy drapes.

Surgical landmarks are identified and marked on the skin, including the superior border of the pes anserine tendons. This is palpable in most patients, but when not palpable, the middle of the tibial tubercle can be used to estimate its location. An oblique, skin-fold incision is used for graft harvest. This incision parallels the gracilis tendon. This incision is approximately 3 cm in length and is centered over the anterior border of the medial collateral ligament (Fig. 38-1). This is generally 4 to 5 cm medial to the tibial tubercle. Both arthroscopic portals and the harvest incision are injected with 5 to 10 mL of 1% lidocaine with epinephrine for the hemostatic effect.¹⁰

Fig. 38-1 Preparing the skin incision for graft harvest. An oblique incision allows for easy harvest visualization, tunnel preparation, and a good cosmetic result.

After the skin incision is made and subcutaneous soft tissues mobilized, the intersection of the anterior border of the medial collateral and the superior border of the sartorius fascia is marked with electrocautery (Fig. 38-2). The main advantage of this starting point is that it allows a relatively horizontal femoral tunnel in the coronal plane when the femoral tunnel is drilled through the tibial tunnel. Vertical grafts in the coronal plane have been associated with increased anterior laxity and functional instability. In order to locate this landmark later in the procedure, the electrocautery is used to make a cruciform periosteal incision that exposes enough of the tibia for the tibial tunnel drill site. The hamstring tendons are harvested after identifying the gracilis and semitendinosus tendons. It is our preference to used a closed loop tendon stripper to remove the tendons after freeing all palpable slips to the gastrocnemius and dissecting the tendon off the tibia, including Sharpey fibers.

Fig. 38-2 Cruciform periosteal incision marking the site of tibial tunnel entry point.

Once the tendons are harvested, they are passed to the back table for preparation. The gracilis and semitendinosus tendons are looped on a suture, creating a four-bundled graft (Fig. 38-3). The free ends of the tendons are sutured with #2 Orthocord (Depuy Mitek) in Krackow-type suture configuration. This is performed in the event that secondary fixation is desired in the tibia. Graft length is generally not an issue, but a minimum graft length of 10 cm, once quadrupled, ensures adequate soft tissue for fixation and ingrowth within the tibia and femur. The proximal end is sutured together with #2 Orthocord for a distance of 3 cm. Each individual graft strand is incorporated to ensure the creation of a single proximal mass. This is helpful to facilitate capture of the graft strands with the Rigidfix pins. Once suturing is complete, the graft is sized using a closed graft-sizing block, with 0.5-mm increments. The femoral side should fit snugly through the sizing block. The tibial side of the graft should freely pass through the sizing block. This, with the additional material of the Krackow sutures, often creates a 0.5 to 1.0 mm difference between the tibial and femoral tunnels. Finally, the graft is marked with a surgical pen 30 mm from the femoral tip of the graft. This allows intraarticular visualization of proper graft seating once it is passed. The graft should be kept under 12 to 15 pounds of tension on the back table until inserted.

Fig. 38-3 Graft preparation for Rigidfix. The proximal segment is sutured between strands to ensure capture by the implant.

Simultaneously with graft preparation, the joint is examined and prepared. Two standard arthroscopic portals are established. A diagnostic arthroscopy is performed, treating any additional pathologic lesions that are encountered. The ACL stump and remnant tissue are removed, clearly identifying the femoral and tibial footprints. Once the soft tissue is removed from the notch, the need for a notchplasty is assessed. A minimal notchplasty is performed to identify the over-the-top position, widen the notch to accommodate the graft, and ensure that the graft will not impinge in full extension. Following completion of the notchplasty, the posterior compartment can be more easily visualized to ensure that no meniscal fragments or loose bodies remain.

At this point, the ACL tibial guide is inserted. The ideal position for the guide is on the lateral downslope of the medial tibial spine, approximately 7 mm anterior to the posterior cruciate ligament (PCL).⁵ Once the tip is firmly seated, correct anteroposterior positioning can be confirmed. The tunnel should be directly medial to the midpoint between the insertions of the anterior and posterior horns of the lateral meniscus. The tip of the “bullet” portion of the guide is then placed on the previously marked junction of the anterior border of the medial collateral ligament and the superior border of the pes tendons on the anteromedial flare of the tibia. The tunnel length is noted and should be at least 30 mm in length. The angle on the ACL tibial guide should be set at approximately 55 degrees. Care is taken not to allow the tip of the bullet to move in a superior direction. If this occurs, an excessively flat and short tibial tunnel will be created. This may compromise both graft fixation and positioning of the femoral tunnel using the endoscopic transtibial approach. For this reason, the bullet is positioned within a few millimeters of the superior border of the sartorius fascia.

A guide pin is drilled through the tibial tunnel guide. Once the pin emerges through the tibial footprint, the drill guide is removed and a hemostat is introduced through the medial portal and clamped to the tibial pin. The tibial tunnel is drilled to the predetermined size while maintaining pressure on the guide pin in a posterior direction using the hemostat. This prevents anterior migration of the intraarticular entry point as the drill penetrates into the joint anterior to the pin. The tunnel is then dilated 0.5 mm greater in diameter than the graft, allowing smooth graft passage. The posterior rim of the intraarticular tibial tunnel is inspected. Soft tissue flaps and any bony fragments are débrided to ensure a smooth surface for the graft in order to minimize abrasion at the tunnel entrance into the joint.

The femoral over-the-top guide is selected. A 1-mm posterior wall is desired. Thus the offset guide should be 1 mm larger than half the diameter of the femoral end of the graft. The correct offset guide is placed through the tibial tunnel and positioned in the over-the-top position. The knee should be placed in 65 to 70 degrees of flexion for drilling of the femoral tunnel. If the guide is not easily positioned with the knee in this position, the knee may be slightly extended for guide placement and then gently flexed into desired position.

Graft positioning in the femur plays a large role in successful reconstruction. The goal is to obtain as horizontal a graft as possible in the coronal plane. The femoral guide should be positioned as far from the 12-o’clock position as possible. The ideal positions to aim for are the 10:00 to 10:30 positions (on the clock face) for a right knee or 2:00 to 2:30 position for a left knee (Fig. 38-4). If this cannot be created through the tibial tunnel, the femoral guide can be placed through the medial portal and the tunnel drilled through this approach. If access through the medial portal is required, the knee should be flexed to 110 degrees to ensure adequate bone for the femoral socket. Care should be taken to not allow the guide to slip anteriorly and to ensure the post remains in the over-the-top position. If anterior slippage were to occur, the femoral tunnel would be placed too anterior, drastically increasing the risk of graft failure. The Beath pin is drilled through the femoral guide through both cortices, but not through the skin of the lateral thigh. It should be palpable below the skin, allowing for easy retrieval if necessary. An acorn drill is used to create the femoral socket to a depth of 35 mm. This depth will allow 30 mm of the graft to be seated in the socket. In smaller and female patients, the cortex may be encountered prior to 35 mm. As long as the drill has penetrated 25 mm, we accept that as the depth because two Rigidfix pins can still be placed to engage the graft. Bone and soft tissue debris are removed from the femoral socket and notch.

Fig. 38-4 Identify the femoral point, and make a tunnel through the tibial tunnel when possible, reaching a depth of 30 mm.

(Reproduced with permission by Depuy Mitek.)

The appropriate-sized cannulated femoral rod is then attached to the cross-pin frame from the Rigidfix system and is inserted into the femoral tunnel over the Beath pin. Once it is fully seated, the Beath pin is removed (Fig. 38-5). If working through the tibial tunnel, the next step is to determine the exact position of joint flexion that provides the least resistance to rotation of the cross-pin guide frame. This step is crucial because knee flexion has often changed subtly since the femoral tunnel was drilled. Flexing and extending the knee a few degrees while rotating the guide frame permits the surgeon to determine which position of the knee allows the cross-pin guide frame to rotate freely and aligns the outrigger properly. This facilitates accurate position of the cross-pin sleeves and decreases the chance that one of the cross-pins will not penetrate the graft. If the guide is properly aligned, it should be nearly parallel to the floor.

Fig. 38-5 Position the Depuy Mitek Rigidfix femoral guide. Note that the arm of the guide is positioned lateral to the knee.

(Reproduced with permission by Depuy Mitek.)

A cross-pin sleeve is then placed over the interlocking trocar. The distal hole in the cross-pin frame is drilled first (Fig. 38-6). The interlocking trocar is drilled until the cross-pin sleeve is fully seated in its hole on the cross-pin guide frame. A small mallet is then used to gently tap the drill to disengage the interlocking trocar from the cross-pin sleeve. When it is disengaged, the user will see that the small interlocking pin on the trocar is no longer engaged with the sleeve. At this point, the trocar may be removed by spinning the drill. The observation of good fluid flow from the sleeve indicates central positioning of the pin in the tunnel. This process is repeated through the proximal slot on the outrigger.

Fig. 38-6 Introduce the guide, and then drill the two cannulated sleeves.

(Reproduced with permission of Depuy Mitek.)

At this point, the cross-pin guide frame is disassembled and removed. The arthroscope is removed from the portal and is inserted through the tibial tunnel into the femoral socket to inspect the cross-pin holes. They should be centered in the tunnel. Passage of guide pins or nitinol wires through the sleeves should be preformed to identify their positions (Fig. 38-7). If only one is centered, we accept this as an acceptable fixation and insert both pins. If neither is correctly placed, graft fixation may not be adequate. At this point, an alternative type of femoral fixation should be employed. It is our preference to use a Milagro bioabsorbable screw, 1 mm over the diameter of the femoral socket. The Beath pin is now reinserted through the femoral tunnel and out through the skin on the lateral thigh. The suture on which the graft was looped is then retrieved and used to pass the graft. The cleft between the strands on the graft should be directed from the anterior to the posterior planes so that the Rigidfix pins will pierce all four arms of the quadrupled graft. The graft is then pulled so that the 30-mm pen mark just enters the femoral tunnel. The tunnel was ideally drilled to 35 mm because the tip of the drill is tapered 5 mm, resulting in 30 mm of cylindrical tunnel and complete fill with a 30-mm graft. As previously mentioned, a 25-mm tunnel (20 mm of graft) can be accepted. The Rigidfix outrigger is designed with 11 mm of bone distal to the more caudal pin, allowing excellent fixation through a shorter socket. The first cross-pin is then inserted using the stepped pin insertion rod and tapped with a small mallet until it is fully seated. Increased resistance occurs as the pin is tapped across the graft. The proximal pin is inserted first. Once it is placed, distal tension of approximately 25 pounds is applied across the graft. This ensures that both pins will share the load. The distal pin is then inserted (Fig. 38-8). The cross-pin sleeves are then removed. It is crucial to cycle the knee to take up any creep that might occur. We routinely use 25 cycles from 0 to 90 degrees with 25 pounds of force using a tensioner.

Fig. 38-7 Arthroscopic inspection of the location of the Rigidfix drill holes.

(Reproduced with permission of Depuy Mitek.)

Fig. 38-8 Femoral fixation with two Rigidfix absorbable pins. The proximal pin is inserted first and the distal pin second.

(Reproduced with permission of Depuy Mitek.)

The tibial side is then fixed using the IntraFix sheath and screw device. The technique for this implant is covered in Chapter 47. Once the graft is fixed on the tibial side, it is again examined arthroscopically. It is placed through a range of motion and checked in full extension to ensure a lack of impingement. The graft is probed to visually check proper tension.

TroubleShooting

Pitfalls may develop at any of several steps in the Rigidfix technique.

The most common pitfall arises in placing the sleeves into the femur using the outrigger guide. If the trocar pin and sheath are drilled into the bone but will not advance to seat fully, be sure that you have removed the Beath pin because this prevents advancing the trocar through the guide. If the Beath pin has been removed and the sheath and trocar will not advance, it is likely that you have changed the angle of the joint enough to slightly deform the guide. Oftentimes the trocar tip will slide into the slot in the guide. Unfortunately, it may also slide off the guide, making the hole for the pin too eccentric to effectively capture the graft. If this occurs and the sleeve has good purchase in the femur, as you visualize the hole with the arthroscope inserted through the tibial tunnel into the femoral socket, you can redirect the trocar and sleeve to be more central in the socket. If this cannot be done, you can remove the sleeves and replace them after deepening the socket by 8 mm.

Removal of the femoral rod may sometimes be difficult. When this happens, it is usually caused by the step between the shaft of the rod and the wider portion from the femur catching on the tibial tunnel. Changing the flexion angle of the knee to match the angle when the femoral socket was drilled and pushing the tibia posteriorly will allow the rod to slide out of the joint.

Two key points for optimal functioning of this system are a snug fit (using 0.5 mm tolerance) of the graft and sutures into the femoral socket and cycling the graft under tension multiple times before fixing the tibial side in order to reduce potential creep from this system. Because the graft is not looped over the pins but rather is skewered by them and compressed against the socket walls, there is the potential for excessive creep in the early postoperative course if these steps are not followed.¹¹ In this study, which was done without cycling the graft prior to testing, 5 mm of creep was found in the first 100 cycles, a large contrast to the manufacturer’s data indicating less than 2 mm of creep over 1000 cycles (assuming adherence to these details).

It is important to tension the graft at or near full extension in order to ensure that the knee does come to full extension in the operating room. The combination of these fixation devices, Rigidfix and IntraFix, results in a construct with very little creep under cyclical loading, and fixing the tibial side with 30 degrees of flexion may lead to overtensioning of the graft.

Postsurgical Care

Immediately postoperatively a hinge brace is applied, with the hinges locked in extension. Icing is helpful for control of pain and swelling. We stress early full extension and have the patient keep the brace on, locked in extension at all times during the first week, with the exception of 6 hours per day in a continuous passive motion (CPM) machine. Ambulation with weight bearing as tolerated is allowed from day 1, with the brace locked in extension. We maintain this weight-bearing protocol for a full 4 weeks after surgery but encourage brace removal after week 1, at any time when the patient is not weight bearing for motion and isometric strengthening. The CPM is discontinued after the 1-week postoperative visit, and formal physical therapy is begun. Quadriceps strengthening is carefully progressed under the supervision of the therapist for the first 4 months. Proprioceptive and agility training are delayed to the 4-month mark, and unrestricted return to athletics is permitted when 6 months has passed, if the thigh musculature is fully rehabilitated.

Results

This system for soft tissue ACL reconstruction, Rigidfix and IntraFix, reliably results in stable knees with 3 mm or less increased translation by KT-1000 at maximal manual pull in our early (2-year) follow-up. Endpoints on drawer and Lachman tests are high pitched. This system offers 360 degrees of circumferential fixation, which is radiolucent and absorbable. All these factors make soft tissue ACL reconstruction an attractive and reproducible procedure.