4 Knee Injuries

The anterior cruciate ligament (ACL) is the most frequently completely disrupted ligament in the knee; most of these injuries occur in athletes (Fig. 4-1). More than 100,000 ACL reconstructions are done each year in the United States.

Figure 4-1 Anterior cruciate ligament and anatomic knee structures.

(Redrawn with permission from Miller MD, Howard RF, Planchar KD. Surgical Atlas of Sports Medicine. Philadelphia, 2003, Saunders, p. 74, Fig. 10-3.)

About 80% of sports-related ACL tears are noncontact injuries, occurring during pivoting maneuvers or landing from a jump. Noncontact ACL injuries are more common in females than in males (see section on ACL injuries in female athletes). Only 60,000 individuals with ACL deficiency actually undergo reconstruction annually.

Hewett et al. (2005) in a level II study found that prescreened female athletes with subsequent ACL injury demonstrated increased dynamic knee valgus (Fig. 4-2) and high knee abduction loads on landing from a jump. Knee abduction moments, which directly contribute to lower extremity dynamic valgus and joint knee load, had a sensitivity of 78% and specificity of 73% for predicting future ACL injury. Neuromuscular training has been shown to decrease knee adduction moments at the knee (Hewitt et al. 1996), and this will be addressed at great length in the ensuing chapter.

Figure 4-2 A, Dynamic valgus is defined as the position or motion, measured in three dimensions, of the distal femur toward the distal tibia away from the midline of the body. Dynamic valgus includes the indicated motions and moments. B, In individuals with anterior knee pain the alignment may appear rather straight—no excessive genu valgum or valgus—but there is significant internal rotation of the femors, indicating femoral anteversion. The patellae are pointing toward one another (left). This is accentuated when the individual gets in a flexed position: the femur goes into further adduction and internal rotation (right).

(Redrawn with permission from Hewett TE, Myer GD, Ford KR, Heidt RS Jr, Colosimo AJ, McLean SG, van den Bogert AJ, et al. Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes. Am J Sports Med 33:4, 2005.).

Although the natural history of the ACL-deficient knee has not been clearly defined, it is known that ACL injury often results in long-term problems, such as subsequent meniscal injuries, failure of secondary stabilizers, and development of osteoarthritis (OA).

Although a number of studies have suggested that OA eventually develops in 60% to 90% of individuals with ACL injuries (Beynnon 2005 Part 1, Andersson et al. 2009), a recent systematic review of the literature (Ojestad et al. 2009) concerning OA of the tibiofemoral joint more than 10 years after ACL injury suggests that these estimates are too high. The lack of a universal methodologic radiographic classification made it difficult to draw firm conclusions, but these investigators determined that in the highest-rated studies the reported prevalence of knee OA after isolated ACL injury was between 0% and 13%, and with meniscal injury, it was between 21% and 48% (level II evidence).

Associated meniscal injury is the most commonly cited factor contributing to the development of OA after ACL injury, followed by articular cartilage injuries. A 7-year prospective study of patients with reconstruction of an acute ACL injury found that 66% of those with concomitant meniscectomy developed OA, compared to only 11% of those without meniscal injury (Jomha et al. 1999). Subjective follow-up of 928 patients 5 to 15 years after ACL reconstruction found normal or nearly normal knees in 87% of patients with both menisci present, compared to 63% of those with partial or total meniscectomies (Shelbourne and Gray 2000). Of 54 National Football League players who had meniscectomy or ACL reconstruction or both, those with both procedures had shorter careers (fewer games started, fewer games played, and fewer years in the sport) than those with either procedure alone (Brophy et al. 2009).

Successful reconstruction of the ACL has been proven to improve short-term function and perhaps decrease the risk of subsequent meniscal injury, but it may not decrease the likelihood of OA (Lohmander and Roos 1994), particularly in patients with concomitant meniscal or articular cartilage injuries.

Treatment of ACL Injuries

Nonoperative Treatment (ACL-Deficient Knee)

• Levy and Meier (2003) reported the incidence of subsequent meniscal tears in ACL-deficient knees is 40% at year 1, 60 percent at year 5, and 80% by 10 years after the initial untreated ACL disruption.

• Lohmander and Roos (1994) in a meta-analysis of 33 studies found that the efficacy of ACL reconstruction in retarding the progression of OA was not substantiated. Presence of meniscal injury at time of ACL injury has a high correlation with eventual development of arthritis.

• Some patients who are ACL-deficient, however, have physiologic responses and motor control strategies that allow successful compensation for their ACL absence (copers). Copers are defined patients who have returned to full sports and preinjury activity without instability for at least 1 year.

• Nakayama and Beard have similiarly demonstrated much improved dynamic knee stability and function in patients with ACL deficiency after rehabilitation that included perturbation training. Perturbation training for nonoperatively treated and postoperative ACL reconstruction in addition to traditional strengthening should be advocated.

Despite the success of current ACL reconstruction methods, not all patients require surgical reconstruction. Currently, there are no firm criteria for determining which patients are candidates for ACL reconstruction versus nonoperative management.

Several authors have suggested criteria for nonoperative treatment in ACL tears: Fitzgerald et al. (2000) developed guidelines for selecting appropriate candidates for nonoperative ACL deficiency management (e.g., initiation of perturbation and strengthening program). The primary criteria were no concomitant ligament (e.g., medial collateral ligament) or meniscal damage and a unilateral ACL injury. Other criteria include the following:

1. Timed hop test score of 80% of the uninjured limb

2. Knee Outcome Survey Activities of Daily Living Scale score of 80% or more

3. Global rating of knee function of 60% or more

4. No more than one episode of giving way in the time from injury to testing

The success rate in Fitzgerald’s perturbation ACL rehabilitation group was 92% (11/12 patients). The likelihood ratio calculated for this study suggested patients would be five times more likely to successfully return to high-level physical activity if they receive the perturbation training than if they receive only a standard ACL rehabilitation strength training program.

Moksnes et al. in a level Ib study (2008) found that 70% of patients classified as potential noncopers in Fitzgerald’s original screening examination were true copers after 1 year of nonoperative treatment.

1. These authors’ observation was that the development of knee function in subjects with nonoperatively treated ACL injuries simply took time.

Most reports of successful nonoperative treatment of ACL injuries come from case series (level IV evidence). One prospective cohort study (level II evidence) of 100 consecutive patients with nonoperatively treated (early activity modification and neuromuscular knee rehabilitation) ACL injuries found that at 15-year followup 68% had asymptomatic knees (Neuman et al. 2008).

Of four randomized controlled studies comparing nonoperative to operative treatment (level I evidence), one reported no difference in outcomes (Sandberg et al. 1987) and three reported superior results with operative treatment (Andersson et al. 1989 and 1991, Odensten et al. 1984).

Although age of more than 40 years has been considered a relative indication for nonoperative treatment, several studies have reported results in older patients similar to those in younger patients, and age alone is not an absolute indicator for nonoperative treatment. Many individuals aged 40 years and older remain athletically active and are not willing to accept the limitations knee instability places on their activities.

Operative ACL Reconstruction

ACL reconstruction is almost universally recommended for patients with high-risk lifestyles that require heavy work or who participate in certain sports or recreational activities. Other indications for ACL reconstruction include repeated episodes of giving way despite rehabilitation, meniscal tears, severe injuries to other knee ligaments, generalized ligamentous laxity, and recurrent instability with activities of daily living (Beynnon et al. 2005 Part 1). Once operative reconstruction is chosen, a number of controversial areas must be considered: timing of surgery; choice of graft, autograft, or allograft; one- or two-bundle technique; fixation method; and rehabilitation protocol (accelerated or nonaccelerated).

A study of National Basketball Association players with ACL injuries and subsequent reconstruction by sports medicine physicians found that 22% did not return to competition and 44% of those who did return had decreases in their levels of performance despite reconstruction (Busfield et al. 2009, level IV evidence).

Timing of surgery. Because many patients had difficulty regaining full knee motion after acute or early reconstruction, delayed reconstruction has been suggested to minimize the possibility of postoperative arthrofibrosis. Good results have been reported after both acute and delayed reconstruction, mostly in retrospective case series. A prospective study compared outcomes in patients who had ACL reconstruction at four time points after injury (Hunter et al. 1996): within 48 hours, between 3 and 7 days, between 1 and 3 weeks, and more than 3 weeks. They found that restoration of knee motion and ACL integrity after ACL reconstruction was independent of the timing of surgery. Shelbourne and Patel (1995) suggested that the timing of ACL surgery should not be based on absolute time limits from injury. They reported that patients who had obtained an excellent range of motion (ROM), little swelling, good leg control, and an excellent mental state before surgery generally had good outcomes, regardless of the timing of surgery. Mayr et al. (2004) confirmed these observations in a retrospective review of 223 patients with ACL reconstructions: 70% of patients with a swollen, inflamed knee at the time of undergoing ACL reconstruction developed postoperative arthrofibrosis. It appears that the timing of reconstruction is not as important as the condition of the knee before surgery: full ROM, minimal effusion, and minimal pain are required (Beynnon et al. 2005, Part 1).

Graft choice. Bone-patellar tendon-bone (BPTB) autografts (Fig. 4-3) have been historically considered the “gold standard” for ACL reconstructions, although good outcomes have been reported with other graft choices, particularly hamstring grafts (Fig. 4-4 A–H). A number of studies have compared BPTB grafts with four-strand hamstring grafts, with most reporting no significant difference in functional outcomes, although difficulty with kneeling was more commonly reported by those with BPTB grafts.

Figure 4-3 A, Bone patellar tendon bone graft harvest. The tendon is exposed, and the paratenon is incised. An appropriately sized graft is measured, and the tendon is incised parallel to its fibers. An oscillating saw is used to remove 25-mm bone blocks from the tibia and the patella. B, An osteotome is used to remove the bone blocks. (Reprinted with permission from Miller MD, Howard RF, Planchar KD. Surgical Atlas of Sports Medicine. Saunders, Philadelphia, 2003, p. 46, Fig. 7-4.) C, Anterior cruciate ligament bone patellar bone graft fixation.

(Reprinted with permission from Miller MD, Howard RF, Planchar KD. Surgical Atlas of Sports Medicine. Saunders, Philadelphia, 2003, p. 57, Fig. 7-14,)

Figure 4-4 A and B, Hamstring graft harvest. Initial dissection beneath sartorial fascia and isolation of gracilis tendon (superior) and semitendinosus tendon (inferior). C, Sutures placed near insertion of each tendon with a whip stitch; harvesting performed with a tendon stripper. D, Passage of the tendon stripper outside of the fascial sling beneath the semimembranous muscle may result in the tendon stripper’s taking an aberrant path into the thigh and causing premature amputation of the semitendinosus graft. E, Muscle is cleared off each tendon with a curette. F, Sutures are placed on the free ends of the graft.G, A fixation device is prepared after the graft has been sized. H, Anterior cruciate ligament graft passage with fixation of the metal EndoButton on the lateral femoral cortex.

(A, B, C, E, F, G Reprinted with permission from Miller MD, Howard RF, Planchar KD. Surgical Atlas of Sports Medicine. Saunders, Philadelphia, 2003, Figs. 7-5A, C, D, 7-8 A, B, C, 7-13. Part D adapted with permission from Brown CH, Sklar JH. Endoscopic anterior cruciate ligament reconstruction using quadrupled hamstring tendons and EndoButton femoral fixation. Tech Orthop 13:285, 1998.)

A meta-analysis by Yunes et al. (2001) found that patients with BPTB grafts had anteroposterior knee laxity values that were closer to normal than did those with four-strand hamstring grafts, and a later meta-analysis by Goldblatt et al. (2005) found that more patients with BPTB grafts had KT-1000 manual-maximum side-to-side laxity differences of less than 3 mm than did those with four-strand hamstring grafts; fewer of those with BPTB grafts had significant flexion loss. Those with hamstring grafts had less patellofemoral crepitance, anterior knee pain, and extension loss.

Autograft versus allograft. Suggested advantages of allografts over autografts include decreased morbidity, preservation of the extensor or flexor mechanisms, decreased operative time, availability of larger grafts, lower incidence of arthrofibrosis, and improved cosmetic result. Disadvantages of allografts include risk of infection, slow or incomplete graft incorporation and remodeling, higher costs, availability, tunnel enlargement, and alteration of the structural properties of the graft by sterilization and storage procedures. Two meta-analyses comparing autografts and allografts found no significant differences in short-term clinical outcomes (Foster et al. 2010, Carey et al. 2009); however, Mehta et al. (2010) found higher revision rates with BPTB allografts than with autografts and higher IKDC (International Knee Documentation Committee) scores in those with autografts.

A prospective comparison (level II evidence) of outcomes of 37 patients with autografts and 47 with allografts found similar clinical outcome scores at 3 to 6 years after surgery (Edgar et al. 2008). A retrospective review of 3126 ACL reconstructions (1777 with autografts and 1349 with allografts) found that the use of an allograft did not increase the risk of infection (less than 1% in both groups); hamstring tendon autografts had a higher frequency of infection than either BPTB autografts or allografts (Barker et al. 2009).

Single-or double-bundle reconstruction. The rationale for two-bundle reconstruction is based on the identification of two distinct ACL bundles: the anteromedial (AM) and the posterolateral (PL) bundle (Fig. 4-5). The femoral insertion sites of both bundles are oriented vertically with the knee in extension, but they become horizontal when the knee is flexed 90 degrees, placing the PL insertion site anterior to the AM insertion site. When the knee is extended, the bundles are parallel; when the knee is flexed, they cross. In flexion, the AM bundle tightens as the PL bundle becomes lax, while in extension the PL bundle tightens and the AM bundle relaxes.

Figure 4-5 A Anterior cruciate ligament (ACL) tear of anteromedial and posterolateral bundles, each from femoral insertion. Preoperative examination demonstrated 2+ Lachman test score and 3+ pivot shift test score.

(Reprinted with permission from Cole B. Surgical Techniques of the Shoulder, Elbow, and Knee in Sports Medicine. Philadelphia: Saunders, 2008. p. 664, Fig. 65-4.)

B The ACL is divided into three bundles based on the tibial attachment: the anteromedial (AM), the intermediate (I), and the posterolateral (PL) bundles. With knee flexion, the posterior fibers loosen and the anteromedial fibers coil around the posterolateral ones.

(Redrawn with permission from Baker CL Jr. The Hughston Clinic Sports Medicine Book. Baltimore: Williams & Wilkins, 1995.)

These observations indicate that each bundle has a unique contribution to knee kinematics at different flexion angles. Cadaver studies have shown that double-bundle reconstructions more closely restore normal knee kinematics (Tsai et al. 2009, Morimoto et al. 2009, Yagi et al. 2002), including a more normal tibiofemoral contact area (Morimoto et al. 2009), than do single-bundle reconstructions. Several prospective, randomized comparisons (level I evidence) of the two techniques have shown superior objective results with double-bundle reconstruction but no significant differences in subjective and functional results (Sastre et al. 2010, Jarvela et al. 2008, Aglietti et al. 2010, Siebold et al. 2008) even in high-level athletes (Streich et al. 2008).

A meta-analysis of the literature (Meredick et al. 2008) found no clinically significant differences in KT-1000 or pivot shift results between double-bundle and single- bundle reconstruction. Other authors have reported significantly more rotational stability after double-bundle reconstruction than after single-bundle procedures (Tsai et al. 2009, Hofbauer et al. 2009, Kondo et al. 2008). The primary disadvantage of double-bundle reconstructions is their complexity and technical difficulty. The creation of multiple tunnels increases the risk of tunnel misplacement and makes revision surgery extremely difficult.

Cited advantages of single-bundle techniques include proven success, less technical difficulty, less tunnel widening, fewer complications, easier revision, lower graft cost when allograft is used, lower implant cost, and shorter surgical time (Prodromos et al. 2008).

Method of fixation. A variety of fixation devices are used for ACL reconstruction, with no consensus as to what is best. Generally, fixation can be classified as interference screw-based, cortical, or cross-pin (Prodromos et al. 2008). Interference screw and cortical fixation can be used in both the femur and the tibia. Interference screw fixation functions by generating frictional holding power between the graft and the bone tunnel wall (Prodromos et al. 2008). Cortical fixation can be direct, compressing the graft against the cortex, or indirect, connecting the graft to the cortex with some sort of interface, often a fabric or metal loop through which the graft is passed. Cross-pinning is a relatively new fixation technique for which advocates cite the advantage of being closer to the tunnel opening than cortical fixation. This advantage, however, has not been proved. A meta-analysis showed that cortical fixation provided more stability than aperture fixation (Prodromos et al. 2005), and a prospective comparison of three fixation devices, including cross-pin fixation, found no statistically or clinically relevant differences in results at 2-year follow-up (Harilainen and Sandelin 2009). All currently used fixation techniques appear to provide adequate stability to allow early aggressive rehabilitation after ACL reconstruction (Hapa and Barber 2009).