➢ Multiligament knee injuries have a strong association with periarticular soft-tissue and neurovascular injuries, which must not be overlooked in the initial evaluation of the patient.
➢ Even though magnetic resonance imaging (MRI) is imperative for a complete evaluation of the damaged ligamentous knee restraints, stress radiography aids in establishing the functional consequence of the MRI findings and may assist in directing reconstruction.
➢ Although cruciate ligament tears are generally reconstructed, a combined repair-reconstruction approach is most useful for collateral ligaments and extra-articular structures, with incorporation of local tissue into the reconstruction whenever possible.
➢ Regardless of the timing and operative technique chosen, patients with multiligament knee injuries are at high risk for complications and long-term disability.
Much attention has been paid to injuries of the anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), lateral collateral ligament and posterolateral corner, and medial collateral ligament and posteromedial corner. When two or more of these are disrupted, the term multiligament injury is often used, and a knee dislocation or substantial subluxation is likely to have occurred1-3.
Multiligament knee injuries can be categorized as high-velocity, low-velocity, and ultra-low-velocity mechanisms. High-velocity injuries result from direct knee impact during motor vehicle accidents, motorcycle collisions, and falls from a height3-9. Low-velocity injuries often occur during sporting events and are more commonly twisting or direct contact injuries9,10. Ultra-low-velocity dislocations have received increased attention recently and occur primarily in obese patients who sustain injuries while doing activities of daily living such as getting out of bed or after a ground-level fall8,11-15. In a series of twenty-three ultra-low-velocity dislocations, Werner et al. noted a mean body mass index of 49.1 kg/m2, and patients were more likely to be female and to sustain an associated neurovascular injury14.
In the classification scheme proposed by Schenck and later modified by Wascher, each knee dislocation is designated with KD (knee dislocation) and is followed by the number of ligaments (or complexes) disrupted (I to IV) and “M” if the posteromedial corner is injured or “L” if the posterolateral corner is injured. The knee dislocation may be further classified to include a type V for periarticular fractures, a “C” modifier for arterial injuries, and an “N” modifier for nerve injuries (Table I)7,16. Importantly, it is possible to have corner incompetence without collateral ligament injuries, and this may manifest with rotational instability but no varus or valgus laxity.
Fortunately, multiligament knee injuries are rare, with an estimated prevalence of 0.02% to 0.2% of all orthopaedic injuries3,17. Compared with an isolated cruciate or collateral ligament injury, intermediate-term and long-term outcome studies have shown a higher rate of complications and varied outcomes with operative and nonoperative management2,3,18,19. This article describes multiligament knee injuries in depth, with a focus on associated injuries, operative management, outcomes, and complications.
The substantial forces required to tear multiple ligaments lead to substantial associated damage, such as injuries to the popliteal artery, peroneal nerve, and other knee soft-tissue structures. Knee dislocations may be associated with trauma to other body systems, particularly in high-velocity subtypes1,5,20-23.
The popliteal artery is at risk for injury during knee dislocation because of its anatomic constraints; it is fixed proximally at the fibrous insertion of the adductor magnus onto the medial femoral epicondyle and is tethered distally by the tendinous arch of the soleus24,25. The reported prevalence of vascular injury associated with knee dislocation varies widely, ranging from 3.3% to 64% depending on injury mechanism, ligamentous injury pattern, and various other patient and injury factors24-28. A recent database study of 8050 knee dislocations estimated the risk of vascular injury to be 3.3%, with patients who are male and twenty to thirty-nine years of age having an increased risk25.
The diagnosis of popliteal artery injury is a subject of controversy, as numerous authors have advocated routine arteriography for any patient with a knee dislocation but authors of more recent studies have questioned the routine use of arteriography and argued that it should be utilized selectively7,24,26-28. Arteriograms are recommended for an ankle-brachial index of <0.8 with a well-perfused foot; any color, temperature, or pulse alterations of the ipsilateral foot; or an expanding hematoma in the popliteal fossa; however, the study should never delay operative intervention of obvious arterial injuries7,26,28. Bedside duplex ultrasonography is an additional modality to rapidly identify vascular lesions and has been demonstrated to be 95% sensitive, 99% specific, and 98% accurate for neck and extremity arterial injuries following penetrating or blunt trauma29. Unfortunately, knee dislocations associated with popliteal artery injury generally have a poor prognosis with high rates of eventual amputation5,14,24,25.
The prevalence of associated neurologic dysfunction with multiligament knee injury or knee dislocation is reported to range from 10% to 40%, with most studies showing a prevalence between 25% and 35%5,30-35. The common peroneal nerve is most commonly injured because of anatomic constraints on its ability to accommodate to traumatic changes in knee position, both proximally at the fibular neck and distally at the intermuscular septum36.
Historically, prognosis for recovery of peroneal nerve function is considered to be poor33,34. Male sex, higher body mass index, and fibular head fractures are risk factors for peroneal nerve injury, but younger age is associated with neurologic recovery36. Other authors have recommended neurolysis when early clinical examination and electromyography (EMG) demonstrate lack of recovery30. EMG and nerve conduction velocities at three weeks after the injury provide information regarding the baseline neurologic function and can be compared with repeat testing at three months to assess for recovery37. Autografting of segmental defects has variable results; grafts of <6 cm demonstrated excellent functional outcomes but grafts of >12 cm had good outcomes in only two (25%) of eight patients31. Posttraumatic peroneal nerve palsy portends substantial morbidity due to the resulting foot drop and likely need for an orthosis5,18,38,39. Foot drop that persists at least one year after knee dislocation can be effectively treated with posterior tibial tendon transfer, which leads to good improvement in motion and dorsiflexion strength40.
In addition to ligamentous injury, the dislocated knee is at risk for injury to other structures, particularly the menisci. With regard to medial meniscal injuries, Chahal et al. reviewed post-injury magnetic resonance imaging (MRI) of twenty-seven high-grade multiligament knee injuries in twenty-two patients and found that nine patients (41%) had medial meniscal injuries41. Werner et al. reviewed MRI studies of fourteen Schenck KD-IIIM knees and eighteen KD-IV knees and found that 43% of KD-IIIM knees and 44% of KD-IV knees had medial meniscal injuries42.
Patients with knee dislocations frequently have associated bodily injuries, particularly those patients with high-velocity subtypes26. In a report on patients with high-velocity knee dislocations, Wascher reported that the rate of life-threatening injuries involving the head, chest, or abdomen was 27%, with a 50% to 60% rate of associated fractures7. Andruszkow et al. reported that severe additional limb injuries and severe chest injuries are more common in polytraumatized patients with knee injuries compared with those without knee injuries43.
MRI is central to characterizing ligament injuries after knee dislocation, as well as identifying extraligamentous or other soft-tissue knee injuries35,44 (Fig. 1). MRI is extremely sensitive for detecting cruciate ligament injury, collateral ligament injury, and injury to the posterolateral and posteromedial corner structures42,44.
Halinen et al. compared MRI with intraoperative findings for forty-four patients with acute multiligament knee injuries45. The authors reported that the accuracy of MRI for detecting medial meniscal tears was 88.6%, with a sensitivity of 80% and a specificity of 91.2%. For lateral meniscal tears, the sensitivity of MRI was 55% and the specificity of MRI was 87.5%. Both the accuracy and sensitivity of MRI were 93.2% for detecting the severity of ACL tears and 86.4% for detecting the severity of medial collateral ligament tears. Nicandri et al. evaluated forty-two patients with multiligament knee injuries and found that MRI demonstrating grade-I PCL tears may predict stability on posterior drawer examination at the time of surgery, but MRI demonstrating grade-II or III PCL tears may not predict clinical instability or the need for PCL reconstruction at the time of surgery46.
Although MRI is highly sensitive for detecting ligament damage, it cannot demonstrate the functional consequences of ligament injuries as it is a static study. Stress radiography offers the ability to obtain dynamic images to supplement MRI, providing an objective representation of the functional laxity resulting from ligament deficiency47.
Gwathmey et al. reviewed twenty-seven knee injuries and found that varus stress radiography corresponded with MRI findings for posterolateral corner injury and used the degree of opening on varus stress radiography to aid in the decision for stabilization47. Jacobson and Chi recommended a similar protocol for medial-sided injuries, advocating for valgus stress radiographs to assess the functional consequence of posteromedial corner injury identified on MRI48. Sekiya et al. performed stress radiography on ten pairs of cadaver knees with an intact PCL and posterolateral corner and then performed it again after sequentially resecting the PCL followed by the posterolateral corner. The authors found that >10 mm of posterior tibial translation on stress radiography correlated with the presence of a posterolateral corner injury in addition to complete disruption of the PCL49. In addition to preoperative assessment, stress radiography can be utilized to quantify and to follow postoperative stability50. At our institution, intraoperative stress radiographs are obtained on all patients to direct the reconstruction and repair of soft tissues.
Operative Management Compared with Nonoperative Management
Peskun and Whelan performed a literature review including 916 patients from thirty-five studies to compare outcomes of operatively treated multiligament knee injuries with those of nonoperatively treated multiligament knee injuries51. They demonstrated that operative management results in improved functional outcome, instability, contracture, return to activity, return to work, and return to sport. At our institution, every patient with a multiligament knee injury is offered surgery, although some patients with ultra-low-velocity injuries are definitively managed with external fixation because of the increased complications noted with reconstruction in this cohort14.
Timing of Reconstruction
The treatment of multiligament knee injuries remains controversial. Many surgeons have reported acceptable outcomes in association with acute surgical repair or reconstruction of injured ligaments, although others have advocated for delayed treatment if collateral ligament injuries are present, and still others have reported favorable results when the repair and reconstruction are staged32,50,52-55.
Supporters of acute reconstruction of all injured structures argue that reestablishing the central axis of knee motion by reconstructing the cruciate ligaments is vital for proper restoration of the anatomic relationships of the collateral ligament and corner structures6. Mook et al. systematically reviewed twenty-four studies including 244 patients who were managed with acute reconstruction of injured structures55. The authors found that acute treatment was associated with increased residual anterior knee instability when compared with delayed reconstruction (odds ratio, 2.58; p = 0.018). Significantly more patients who were managed acutely were found to have flexion deficits of ≥10° when compared with those who were managed in a delayed fashion (odds ratio, 5.18; p = 0.004). Consequently, additional treatment for joint stiffness was significantly more likely (p < 0.001) in association with acute treatment (17%) compared with delayed treatment (0%)55.
Staged Acute Reconstruction
Staged reconstruction involves acute surgery (within two weeks) of the extra-articular injuries, followed by staged reconstruction of the injured cruciate(s) once full knee motion has been restored50. Early repair and/or reconstruction of the collateral ligament complexes is important, as delay past this time point causes normal anatomy to be obscured by scar tissue, rendering repair difficult and jeopardizing vital neurovascular structures6. Existing literature supports this approach, as staged management had the highest percentage of excellent or good outcomes in KD-IIIM and KD-IIIL knees in a recent systematic review56. Mook et al. reported similar findings, with staged treatments yielding the highest percentage of excellent and good outcomes and the lowest likelihood of requiring additional procedures for joint stiffness55.
Delaying reconstruction offers the advantage of obtaining better preoperative knee range of motion and avoiding unnecessary reconstruction or repair of extra-articular ligaments that may heal with sufficient stability21,22. Systematic reviews have demonstrated moderate results for chronic or delayed reconstruction55,56. Reported outcomes of delayed reconstruction are challenging to interpret, as ligamentous reconstruction in patients with more severe injury patterns, substantial soft-tissue injuries, or concomitant nonorthopaedic injuries may be delayed out of necessity, resulting in worse-than-expected outcomes for the chronic subgroups22.
Repair Compared with Reconstruction
The primary repair of tears of the cruciate ligaments is associated with increased failure rates and poorer results. However, the corners are thought to have higher healing capacity and thus are more amenable to acute repair8,9,21,53,57. Some of the initial reconstructive techniques of these extra-articular structures focused on isolated repair of damaged tissue, but high failure rates and residual laxity have led many to augment with a graft reconstruction57-60.
After medial collateral ligament and posteromedial corner injuries, for example, Stannard et al. noted a failure rate of 20% (five of twenty-five) after repair and only 4% (two of forty-eight) after autograft reconstruction59. However, a systematic review in 2009 by Kovachevich et al. found satisfactory results with either repair or reconstruction of the medial collateral ligament in multiligament knee injuries58. For posterolateral corner injuries, Levy et al. demonstrated a failure rate of 40% (four of ten) after repairs and 6% (one of eighteen) after reconstruction (p = 0.04)60. Similarly, in 2005, Stannard et al. noted a failure rate of 37% (thirteen of thirty-five) after repair and 9% (two of twenty-two) after reconstruction57.
Currently, a combined repair and reconstruction approach to extra-articular structures is utilized on the basis of each unique injury characteristic1-3,6,9,19,22,61,62. In general, soft-tissue repair is not considered an option after approximately three weeks post-injury because of the lack of defined soft-tissue planes6,8. Additionally, reconstruction has gained acceptance due to improvements in allograft processing and thus an increase in the availability of high-quality soft-tissue grafts1,2,21,22. To our knowledge, no large series has demonstrated a higher revision rate compared with that after allograft reconstruction3,14.
Repair techniques for extra-articular structures employ nonabsorbable sutures, suture anchors, and staples2,6 (Fig. 2). Repair is accomplished sequentially from deep to superficial with the goal of restoring normal anatomy. Damaged native tissue may also be sewn side-to-side to grafts to augment reconstructions. Osseous avulsions are repaired with screw or screw-and-washer constructs.
There is substantial controversy over the optimal reconstructive techniques, and no gold standard exists3,9,21. Because of the relative rarity of these events and the unique characteristics of each injury pattern, prospective randomized trials are unlikely to transpire or yield useful information. However, all require a diagnostic arthroscopy to avoid neglecting intra-articular pathology (Fig. 3). In addition to meniscal and cartilage injuries, arthroscopy may allow the surgeon to visualize medial collateral ligament fibers and popliteal tendon injuries and to evaluate the competence of the cruciate ligaments.
For multiligament knee injuries, to our knowledge, there have been no studies that examined single bundle compared with double bundle, transtibial compared with anatomic, interference compared with suspensory femoral fixation, allograft compared with autograft, or hamstring compared with bone-patellar tendon-bone autograft constructs. Thus, the selected arthroscopic surgical technique is typically surgeon-dependent2. Many surgeons preferentially use autograft tissue for cruciate reconstruction and rely on allograft tissue for the injured extra-articular components, although there is no consensus regarding this practice1,6,21. At our institution, ACL reconstruction using independent femoral drilling is our preferred technique, although transtibial and two-incision techniques are also acceptable.
Tibial inlay and transtibial reconstructions are the two main technique options2,6. As the inlay technique necessitates access to the posterior aspect of the knee, patient positioning may have to be altered. Lateral decubitus positioning allows easy access to the posterior and lateral aspects of the knee, and then the patient may be rolled supine for medial knee exposure. As the PCL graft is intra-articular, some surgeons prefer to use autograft tissue if the ACL is not being reconstructed concomitantly6. Patellar tendon allograft or autograft, Achilles allograft, and hamstring autograft or allograft have all been described2,6. However, there is no graft or technique gold standard. In forty-five single-bundle and forty-five double-bundle PCL reconstructions in the setting of multiligament knee injury, Fanelli and Edson demonstrated no significant differences (p > 0.05) in subjective patient outcomes, laxity, or rate of return to pre-injury level2.
Medial Collateral Ligament and Posteromedial Corner
When possible, repair of medial-sided structures should be strongly considered, with graft augmentation as needed2,6. Suture anchors can be used to reattach an avulsed medial collateral ligament, but midsubstance disruptions are often not amenable to primary repair and require reconstruction (Fig. 4). The majority of posteromedial corner reconstructive options focus on the medial collateral ligament, although some authors stress the importance of reestablishing the triad of the medial collateral ligament, posterior oblique ligament, and semitendinosus59. Earlier techniques described using semitendinosus autograft by maintaining the tibial insertion and attaching the proximal portion of the tendon at the femoral origin63,64. Although this places the tibial insertion of the graft too anterior, it is an efficient and cost-effective way to reconstruct the medial collateral ligament65,66. Marx and Hetsroni advocated an Achilles allograft with bone plug placement and interference screw fixation at the femoral origin and a 4.5-mm cortical screw and spiked washer at the tibial insertion67. At our institution, the medial collateral ligament is reconstructed utilizing a modified Bosworth technique with screw-and-washer fixation at the femoral and tibial attachment sites. Regardless of the reconstructive technique, anatomic and isometric positioning of the graft is important and must be tested intraoperatively during knee range of motion6,67.
Lateral Collateral Ligament and Posterolateral Corner
Similar to posteromedial corner reconstruction, several techniques have been developed for posterolateral corner reconstruction. They are roughly divided into anatomic and non-anatomic, with anatomic currently being the preferred option6,68-70. Autograft tissue (split biceps tendon, biceps tendon, semitendinosus) and allograft tissue (Achilles tendon, bone-patellar tendon-bone, tibialis anterior) are both viable options1. Additionally, some surgeons utilize a capsular shift to further stabilize the posterolateral corner2,9,22. Regardless of the technique, the peroneal nerve must be isolated and must be protected throughout the case.
The structures of the posterolateral corner are particularly amenable to repair, which should be considered when possible1,2,6. Osseous injuries such as an arcuate complex avulsion should be reattached to the proximal part of the fibula. Disruption of the proximal tibiofibular joint should not be overlooked. As repair and reconstructive techniques rely on anchoring to the fibular head, this instability must be addressed, commonly with a screw placed across the joint.
Regardless of the surgical technique, postoperative rehabilitation plays an important role in the final outcome in knees with multiligament injuries2,3,6,8,9,22,55,71,72. A balance must be found between immobilizing the knee to allow healing of the soft tissue and early motion to avoid arthrofibrosis6. There is no uniform rehabilitation protocol used among all surgeons, and each plan is tailored for each patient on the basis of the injured structures8.
In general, patients are fitted with a hinged knee brace to be worn for the first six weeks at all times. For the first two to six weeks postoperatively, patients are kept either non-weight-bearing or toe-touch weight-bearing. For some therapy protocols, therapy begins on postoperative day one with prone active and passive range of motion. If the PCL was reconstructed, then active quadriceps exercises are avoided for six weeks. Other protocols, such as the one established by Fanelli and Edson, recommend immobilization in extension for three weeks54. For athletes, straight-ahead running is gradually allowed, followed by cutting and then sports-specific therapy, and full release to sport participation around nine months2.
When assessing the outcomes of multiligament knee injuries, one must keep in mind that most studies include a variety of knee injury patterns19. Thus, it is difficult to accurately predict the outcome of a patient with a particular ligament injury. In a Level-IV study, Fanelli and Edson evaluated the outcomes of forty patients who underwent ligament reconstruction for global laxity (ACL, PCL, posterolateral corner, and posteromedial corner injuries)2. With a follow-up rate of 70% (twenty-eight of forty) at two to eighteen years, the mean knee ligament rating scale scores were 83.8 points for the Lysholm scale (of a possible 100 points), 4 points for the Tegner scale (of a possible 10 points), and 79.3 points for the Hospital for Special Surgery score (of a possible 100 points). Only 59.3% of patients were able to return to their prior level of activity. Posterolateral rotational stability was corrected or overcorrected in 96.4% (twenty-seven of twenty-eight knees), and KT1000 (MEDMetric) arthrometer side-to-side measurements averaged no more than 2.48 mm in all directions tested.
In 2009, Levy et al. published a systematic review of the current literature on outcomes of surgically and nonsurgically treated multiligament knee injuries21. When compared with patients with nonsurgically treated knees, patients with surgically treated knees had improved rates of return to work and sport activities and increased Lysholm and International Knee Documentation Committee (IKDC) scores. Compared with repair, reconstruction produced superior motion and rate of return to pre-injury activity level and a lower prevalence of failure and positive posterior sag.
In a prospective randomized study of combined ACL and grade-III medial collateral ligament injuries, Halinen et al. evaluated quadriceps strength after either ACL reconstruction plus medial collateral ligament repair or ACL reconstruction and nonoperative treatment of the medial collateral ligament injury71. At fifty-two weeks postoperatively, strength deficits were 30.7% for the group that underwent ACL reconstruction plus medial collateral ligament repair and 20.5% for the group that underwent ACL reconstruction and nonoperative medial collateral ligament treatment. In a separate study with a two-year follow-up, Jenkins et al. demonstrated that quadriceps peak torque had recovered to only 85% of the uninjured side, and only 30% of patients were able to return to the pre-injury occupation or sporting ability73.
Compared with single-ligament injury patterns, operatively treated knees with multiligament injuries have high complication rates1,3,9,50,71,74. When including all complications, rates of 6% to 73.9% are to be expected, and there is a direct correlation of the number of injured ligaments and obesity with the resulting complication rate3,14,75. Complication rates vary widely among studies because series contain patients with very different injuries, types of surgical techniques, rehabilitation protocols, and definitions of complications. Postoperative infections have been reported to occur in 0% to 17.4% of surgical procedures to treat knees with multiligament injuries3,14,62. Risk factors include increased surgical time, prolonged tourniquet use, introduction of foreign material, hematoma formation, and, of particular concern in the ultra-low-velocity category, large dead space, obesity, and diabetes mellitus3,14.
Arthrofibrosis is common, requires surgical treatment in 29% of patients, and is more common after more severe injuries, medial-sided repair or reconstruction, and acute surgery2,3,8,19-21,55,67,76. Cook et al. noted that injuries that included more than two ligaments and acute surgery increased the risk of stiffness requiring a manipulation under anesthesia3. Delayed reconstruction after two to three weeks has been advocated to decrease the risk of this occurrence2. In ACL-deficient knees, one option is to reconstruct or repair the other injured structures first and then to address the ACL in three to six weeks once acceptable range of motion has been established2,77,78.
Because of considerable heterogeneity of injuries and treatments, it is difficult to establish normative values for residual laxity and failure rates. Of thirty-five patients who underwent at least ACL and PCL reconstructions, Fanelli and Edson noted that 54% (nineteen of thirty-five) of the knees had abnormal posterior drawers and/or tibial step-off2. For posteromedial corner injuries, Stannard et al. noted a failure rate of 20% (five of twenty-five) after repair and only 4% (two of forty-eight) after autograft reconstruction59. In a separate study by Stannard et al. on posterolateral corner injuries in fifty-seven knees, there was a 37% repair failure rate compared with a 9% reconstruction failure rate57. With an overall revision rate of 9% (twelve of 133), Cook et al. demonstrated a substantially higher revision rate after KD-IV injuries compared with all others3. Compared with isolated repairs, it appears that reconstruction may be able to provide lower failure and revision rates, but some publications have questioned this idea3,58.
Posttraumatic osteoarthritis is reported in 29.6% to 53% of knees due to the acute cartilage insult as well as nonphysiological loading from residual instability2,6,9. In a retrospective study of eighty-nine patients at a mean follow-up of 8.2 years, Richter et al. utilized the Jager and Wirth scoring system and demonstrated that the degree of osteoarthritis is correlated with the presence of collateral ligament ruptures and the degree of residual instability9.
Multiligament knee injuries represent a serious insult to the complicated stabilizing soft-tissue structures of the knee. As such, the initial management, surgical repair or reconstruction, and postoperative rehabilitation should be directed by a team of experienced professionals. Because of the variety of injury patterns and lack of standardized management strategies, it is difficult to accurately predict outcomes of these injuries or to direct treatment (Table II). However, patients should be counseled with regard to the high rate of complications and possible long-term consequences whether treated surgically or nonoperatively.
Source of Funding: There was no external funding source.
Investigation performed at the Department of Orthopaedic Surgery, University of Virginia Health System, Charlottesville, Virginia
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated