Fig. 8.1
Mean center point of the femoral insertion of the anterior cruciate ligament (After Piefer et al. [9])
On the tibia, the footprint is situated approximately 10 mm posterior to the anterior border of the tibia and is roughly triangular, with the apex posterior, extending anteroposteriorly for between 15 and 19 mm [8]. The mediolateral extent of the tibial footprint is limited by the medial and lateral tibial spines. At the tibia, the anteromedial bundle lies adjacent to the anterior horn of the lateral meniscus, directly anterior to the posterolateral bundle [8]. Fibers of the ACL insertion fan out under the intermeniscal ligament and some may blend with the posterior horn of the lateral meniscus [11].
The ACL ranges in length from 22 to 41 mm, depending on the patient, and have an irregular shape in cross-section, which changes by position. Overall, the ACL becomes broader from proximal to distal, although the narrowest point is slightly distal to the femoral insertion [11]. As it approaches the tibial insertion, the ligament becomes broader; the cross sectional area of the ligament is one third greater at the tibial insertion than at the femur [12]. While the bundles are parallel to one another in extension, in flexion the AM bundle spirals around the PL bundle.
The primary roles of the ACL are to resist tibiofemoral rotation and anteroposterior translation. The two bundles contribute differently to these two roles depending on the degree of knee flexion. At full extension, the AM bundle is vertical within the notch and has therefore not been considered to provide a great deal of rotational constraint; by contrast, the PL bundle is more transverse and so would be likely to provide more rotational stability [13]. In fact, the evidence for either bundle providing a clinically meaningful degree of rotational stability is limited: while cadaveric studies have demonstrated statistically significant increases in rotational laxity when one or other of the bundles are cut, neither appears to produce a clinically detectable effect (and even cutting the entire ACL produces only 4° of additional internal rotation) [5]. More recent work has highlighted the role of the anterolateral ligament (ALL), together with extra-articular structures such as the iliotibial band, in conferring rotational stability [14]. In terms of anteroposterior stability, both the AM and PL bundles are important, but the AM has a more important role in the flexed knee and the PL is more important in extension [5].
The ACL is innervated from a branch of the posterior tibial nerve. Nerve fibers within the ACL are associated with mechanoreceptors which are important in giving the ACL its proprioceptive function. It takes its vascular supply from the medial geniculate artery, with the blood supply predominantly entering the ligament proximally in the intercondylar notch; as a result, the proximal part of the ACL has a more comprehensive vascular supply than the distal part [11].
8.3 Mechanisms and Risk Factors for ACL Injury
Around 70 % of cases of ACL rupture result from non-contact injuries, and the predominant mechanism of injury appears to be a valgus and rotational force exerted on a part-flexed knee [15]. As a result, sports which exert these pivoting forces across the knee are those which carry the highest risk of ACL injury: a recent meta-analysis of ACL injuries amongst American high school athletes demonstrated that the sports with the highest risks of ACL injury were basketball (other studies have found a similarly high risk in netball [16]), association football, American football and lacrosse [17]. Similarly, the new UK National Ligament Registry reports that 42 % of cases resulted from association football, 13 % from Rugby football and 8 % from skiing [18]. While ACL injuries are commoner in men (due to their higher level of participation in sport) [1], women have a higher rate of injury per sporting episode [17]. The reasons for this are poorly understood but are likely to be due to differences in limb alignment and a lower rate of muscle development in women during puberty [19]. Other non-sex-specific risk-factors include environmental factors (injury is more common in dry conditions or on artificial turf), anatomical factors (a narrow intercondylar notch and a greater tibial slope predispose to injury, as do generalized or knee-specific joint laxity) and neuromuscular/biomechanical factors such as a reduced range of motion of the hips [20]. Understanding of these risk-factors has led to increased research into the use of neuromuscular training to prevent ACL injury. Amongst such studies, the largest is a randomized study of 4,564 footballers in Sweden, where a neuromuscular warm-up program led to a statistically significant decrease the relative risk of cruciate ligament injury by 64 % compared to controls, although the absolute difference in rate of injury was non-significant [21].
8.4 Clinical and Radiological Examination in ACL Injury
The process of clinical and radiological examination of the knee is covered in detail in Chaps. 1 and 2 and it is beyond the scope of this chapter to provide more than a broad overview of these topics.
8.4.1 History and Clinical Examination
The diagnosis of ACL injury is often made on the basis of the mechanism of injury. As described in Sect. 8.3, ACL injury is commonly the result of non-contact twisting injuries during sport. Patients will often describe an immediate large haemarthrosis and will normally not be able to continue playing. Sometimes, an audible ‘pop’ is reported [22]. The acute injury is painful, probably as a result of the haemarthrosis and associated bone bruising, but this pain will subside over the early post-injury period.
Examination in the acute setting is hampered by the haemarthrosis, which may mask knee instability [22]. Three main clinical tests have been described for the assessment of ACL injury: the Lachmann test, the Pivot Shift test and the Anterior Drawer. The method of performing these tests are covered in Chap. 1; of the three, the Lachmann test has the highest intra-rater and inter-rater reliability, with the inter-rater reliability being highest when it is performed in the prone position [23]. The Lachmann test is also the test with the highest sensitivity for detecting complete ACL rupture when measured against magnetic resonance imaging (MRI) (with a sensitivity of 96 % in awake examination), but pivot-shift is also highly accurate (95 %). The accuracy of both is much lower in detecting partial tears, and both are made more sensitive by examining the patient under anaesthetic [24].
8.4.2 Imaging Examination
In most cases of acute rupture of the ACL, the diagnosis will be made on the basis of clinical findings [25]. In practice, however, most patients will undergo a plain radiograph and an MRI before surgery, to confirm the diagnosis and to exclude other coexisting pathologies within the knee joint.
Radiographs will often demonstrate a haemarthrosis, may show other pathology such as associated osteochondral defects and may demonstrate a Segond fracture (an avulsion fracture from the lateral border of the tibia, which may represent an avulsion of anterolateral capsule or the anterolateral ligament [26]) in up to 13 % of cases [27].
Characteristic MRI findings following ACL injury include discontinuity of the fibers of the ligament, with increased signal on T2 weighted images. Secondary signs include bucking of the PCL (Fig. 8.2), uncovering of the posterior horn of the lateral meniscus and anterior displacement of the tibia (which may also be seen in plain radiographs) [27]. Acute imaging will often demonstrate a characteristic pattern of bone bruising. Most commonly, this represents the ‘pivot shift’ mechanism of injury with bruising seen in the lateral femoral condyle and the posterolateral tibial plateau, although other patterns of bone bruising have been described [28]. Finally, other injuries to ligament, cartilage or meniscus, may be present. A recent meta-analysis of diagnostic accuracy studies of MRI in traumatic knee pathology has reported a pooled sensitivity and specificity for MRI in diagnosing ACL injury at 87 % and 93 % respectively [29], although some studies suggest both figures are as high as 100 % [30]. More details of imaging protocols and findings are given in Chap. 2.
Fig. 8.2
Sagittal MRI image demonstrating absence of the ACL in a chronic tear with bucking of the PCL
8.5 Consequences and Natural History of ACL Injury
8.5.1 Clinical Outcomes
As covered in Sect. 8.1, the ACL is an important driver of knee function and rupture of the ACL leads to a change in the kinematics of the knee during activities of daily living [31]. However, changes in kinematics can be tolerated; the outcomes of interest after ACL rupture are pain and loss of function as measured by patient-reported outcome measures [32], failure to return to sport, further injury and progression to osteoarthritis.
Studies examining the outcomes of conservatively treated ACL rupture are hampered by the selection bias resulting from the removal from study of those treated surgically (which may logically be the more symptomatic or those in higher-demand individuals). To our knowledge, there is only one randomized, controlled trial comparing non-operative to operative management for ACL rupture [33], and the evidence base for this section includes this trial and a small number of series of conservatively treated patients [34].
Frobell et al. randomized 121 patients to receive either early reconstruction or initial conservative management with the option of later reconstruction if symptomatic, publishing 5-year results in 2013 [33]. Half of those in the delayed group opted to undergo later reconstruction; results were presented both on an intention to treat basis and as treated. While knee stability to examination (Lachman and pivot shift tests) was significantly superior in those who underwent reconstruction, patient-reported outcome knee injury and osteoarthritis outcome score (KOOS) was similar. The incidence of radiographic osteoarthritis (OA) was similar in both groups on both analyses and the rate of subsequent meniscus surgery again did not vary between groups. Taking the cohort of 29 patients who were treated with rehabilitation alone, at 5 years, while only one had a normal Lachman test, the overall KOOS score was good at 82/100 and 3/26 (12 %) had radiographic evidence of OA. In a separate study of the same patients, 3 years postinjury, found that muscle strength and performance was similar in patients who had undergone reconstruction compared to those who were treated non-operatively [35].
Kostogiannis et al. in 2007 published the 15-year results of 100 patients with ACL rupture treated non-operatively [36]. Patients were actively persuaded not to undergo reconstruction primarily and all underwent arthroscopic examination at 10 days to confirm the diagnosis and treat concomitant injuries (which were present in 85 %). Reconstruction was only performed in patients who had unacceptable knee function; patients were excluded from the study if they sustained a contralateral tear (six patients) or were lost to follow-up (six patients). Of the remaining 88 patients, 21 underwent delayed reconstruction, leaving 67 available for analysis at 15 years. Unreconstructed patients displayed a significant decrease in Tegner activity scale but in many cases returned to competitive sports and attained good Lysholm, International Knee Documentation Committee (IKDC) and Knee Osteoarthritis Outcome Scores (KOOS) at least in early follow-up (1–3 years). In spite of this, 49/67 patients had good or excellent Lysholm scores at 15 years. No information is given about radiographic progression of arthritis.
A reduction in activity level following ACL rupture in conservatively-treated patients was observed in the systematic review of Muaidi et al. Fifteen studies met their inclusion criteria of which five were included in a meta-analysis of outcomes with respect to activity level [37]. They report an approximate reduction in level of activity of 21 % following ACL injury. However, good patient reported outcomes were achieved in the short term after injury in conservatively managed patients. Again, it should be re-iterated that in most studies there is inherent selection bias due to removal of patients (who may be the most symptomatic) who elect to undergo reconstruction.
8.5.2 Progression to Osteoarthritis
Long term studies of radiographs following ACL rupture have demonstrated OA in between 24 % and 86 % of cases [34]. Lohmander et al. interviewed 84 female footballers a mean of 12 years following rupture of the ACL, 62 % of whom had undergone reconstruction [38]. Half had never returned to competitive football following their injury, and only 8 % participated at the time of interview. KOOS scores were significantly worse in ACL-injured subjects compared with a reference group of un-injured footballers. Radiographic analysis (performed in 67 subjects) demonstrated half (34 knees) to have tibiofemoral or patellofemoral OA in the affected knee compared to only 8 % with radiographic evidence of OA in the contralateral knee. Neither symptoms nor radiographic changes were affected by whether or not reconstruction was performed.
However, the systematic review of Øiestad et al. has suggested that we may have overestimated the incidence of OA following ACL rupture [39]. Their study included 31 studies, each having a minimum of 10 years’ follow-up with radiographic analysis. For isolated injury to the ACL, between 0 % and 13 % of patients had radiographic evidence of OA, rising to between 21 % and 48 % of patients with associated meniscal injuries. Reasons for the overestimate of OA in previous studies may include the inclusion of patients with associated injuries or selection bias related to symptomatic patients returning to follow-up appointments [34].
The presence of associated meniscal injuries appears to be important in determining whether patients develop later OA. Hart et al. performed single-photon emission computed tomography (SPECT) scans on a cohort of 31 patients, 10 years following ACL reconstruction [40]. Only one of the 16 patients with an isolated ACL rupture showed any evidence of knee degeneration on SPECT, while five of the 15 who had undergone meniscectomy did, a significant difference. The systematic review of van Meer et al., while describing the overall quality of the evidence as being poor, found reasonable evidence of the importance of medial but not lateral meniscal injuries in predicting the development of OA. They found no evidence that increasing time to surgical reconstruction increased the risk of OA in the longer term [41].
Overall, the literature supports conservative management as a viable treatment option in the ACL injured individual. In practice, the decision to offer surgical treatment will depend on the patient, their symptoms and their aspirations. Some patients will be able to return to full sports in spite of their injury while some will be happy to adapt to a lower level of function [42]. Several authors have attempted to quantify these factors; in Kostogiannis’ study, those with a positive pivot at three months were more likely to go on to reconstruction [36], while Fithian et al. defined patients as being at low, moderate or high risk of a poor result following ACL injury on the basis of whether they participated in high level sport and the extent of their arthrometer-measured knee laxity [43], offering early surgery to those deemed high risk. On the basis of the literature, in all but high level sportspeople or those with clearly symptomatic laxity preventing activities of daily living, it appears that initial conservative management with a structured rehabilitation program is a reasonable treatment strategy, with reconstruction being offered to those who are unable to return to their desired level of activity over the months following injury [44].
8.6 Surgical Management of ACL Injuries
Arthroscopic reconstruction of the ACL has been the gold standard for surgical management since the 1980s. Reparative procedures have largely been abandoned as they have a significant rate of re-rupture [45]. Similar principles are followed in all types of ACL reconstruction: a graft is secured in transosseous tunnels placed to recreate the functional anatomy of the native ACL. However, there remains controversy over several aspects of surgical management, including choice of graft, positioning of tunnels, choice of fixation and the use of single- or double-bundle techniques [46]. The aim of this section is to describe the evidence base on which these choices can be made; the techniques themselves are described in detail elsewhere and this is not intended to be a practical guide for those performing the surgery.
8.6.1 Choice of Graft
In both the USA and Europe, most ACL reconstructions are performed using autograft, mainly bone-patellar tendon-bone (BTB) or hamstrings (HS; semitendinosus +/− gracilis tendons) [18, 47, 48]. Allograft and synthetic grafts are rarely used in the primary situation but are more frequently used in revision surgery.
Both BTB and HS grafts are widely used and each has their advantages and disadvantages. Disadvantages to each largely concern donor site morbidity – anterior knee pain and a rare incidence of patellar fracture in BTB grafts, and hamstring weakness in when semitendinosus and gracilis are used [49, 50]. In the patellar tendon, bone blocks are present at each end of the graft which confer theoretical advantages in terms of primary fixation [51]; the size of hamstring grafts is more variable than the size of patellar tendon grafts and small grafts can predispose to failure [52]; hamstrings may provide biomechanical superiority and allow double-bundle reconstructions [53], while BTB grafts may be stiffer, conferring greater knee stability [54].
A 2011 Cochrane review examining this topic reported that, allowing for a lack of long term studies, there were no significant differences in patient-reported outcome between the two techniques [54]; however, they report greater static stability when BTB was used, at the cost of a higher rate of anterior knee pain and loss of range of motion compared to HS. Since then, there have been a number of good-quality studies that have reported, including four randomized controlled trials (RCTs). Sajovic et al. report an RCT comparing 32 patients with BTB grafts to 32 patients with HS grafts at 11 years. They report similar rates of graft failure, radiographic changes and patient-reported outcome, with a statistically significantly higher rate of positive pivot shift in the BTB group [49]. Barrett et al. analyzed a series of 1,131 ACL reconstructions and reported that BTB grafts had a lower failure rate amongst patients under the age of 25 years compared to allograft or hamstrings. This is supported by data from the Norwegian registry which suggests a lower revision rate following BTB grafts compared to hamstrings [55].
Wipfler et al. randomized 62 patients to receive either BTB or HS reconstruction [56]. No significant difference was reported in terms of patient-reported outcome or knee stability at 9 years following surgery; however, more anterior knee pain was reported in the BTB group. Likewise, the study of Björnsson, with 193 patents randomized to receive BTB or HS reconstruction, reported no difference in patient-reported outcome or knee stability (in terms of Lachmann or arthrometer measurements) at 16 years’ follow-up. Patients in the HS group were more likely to have a normal pivot shift (p = 0.048) and had less anterior knee pain, but differences between the groups were small and only of borderline significance [57]. Kautzner et al. report another RCT comparing BTB with HS reconstruction, with a larger study group (150 patients) but at only 2 years’ follow-up, and including only women [58]. No differences were reported in terms of Lysholm score, knee stability or graft failure; again, there was less anterior knee pain in the HS group. The final RCT, of Mohtadi et al., randomized 330 patients to receive either BTB, single-bundle HS or double-bundle HS [59]. Patient-reported outcome measures were similar for all techniques, but BTB grafts demonstrated superior post-operative static stability.
The traditional technique for HS reconstruction has involved the harvesting of both semitendinosus and gracilis, each being doubled to create a four-strand graft. Recently, attention has been paid to gracilis-sparing ACL reconstruction, where the semitendinosus tendon graft is quadrupled. This has the advantage of maintaining a greater level of hamstring function; however, it is more technically demanding and there are few perceptible differences in clinical outcome compared to traditional techniques [60, 61]. Alternative autografts, including iliotibial band (ITB) and quadriceps tendon, have been proposed and appear to have similar long-term results compared with more frequently used grafts [62, 63]. Both may be useful alternatives but lack the weight of evidence supporting the use of more traditional grafts.
The use of allograft is attractive in that it eliminates the problem of graft site morbidity. Achilles allografts share the benefits of patellar tendon autografts in that they have an integral bone block, at least at the distal end (Fig. 8.3). The previously reported high rate of failure for allograft may be due to the use of gamma irradiation, which has a deleterious effect on the biomechanics of the tendon, even at low doses [64, 65]. A number of recent systematic reviews have suggested that the results of non-irradiated allograft are similar to autograft [66, 67], although the majority of studies are of older patients and there is some doubt as to the applicability of these results to younger patient groups [68]. Synthetic devices have the potential to eliminate graft site morbidity without the risks associated with the use of allografts; however, previous generations of synthetic ligament had high rates of failure related to non-infective synovitis [69]. These risks seem to be substantially lower in modern devices such as the Ligament Augmentation and Reconstruction System (LARS; Surgical Implants and Devices, Arc-sur-Tille, France), which have encouraging published results, at least in the short to medium term [70]. There is, however, little long-term evidence to support its routine use in the primary situation.
