Introduction

Successful anterior cruciate ligament (ACL) reconstruction is defined by complete return of knee function without instability [1–3]. Numerous in vitro studies have shown that a deficiency of posterolateral corner (PLC) structures results in persistent knee instability and increased load on the grafts used for cruciate ligament reconstruction [4–9]. LaPrade et al. demonstrated that injuries to the PLC can result in a significant increase in the varus load on an ACL graft, thus increasing the risk for graft failure [6]. More recently, Bonanzinga et al. demonstrated that even in the presence of a partial lesion of the PLC, an isolated ACL reconstruction is not able to fully control rotational laxity of the knee [5].

These findings are consistent with data from clinical series with high rates of ACL graft rupture (ranging from 15 to 24%), being attributed to a lack of recognition of the PLC injury [6, 10, 11]. High rates (50–76%) of missed diagnoses of PLC injuries, even after a clinical review of patients by orthopedic surgeons, are reported [11]. Some authors have suggested that the correct diagnosis is frequently challenging owing to the anatomical complexity of this region and difficulty in interpreting MRI findings in the presence of diffuse post-traumatic soft-tissue edema [12–14]. To reduce the rate of missed diagnoses, it is important to have an appropriate index of suspicion for PLC injury [15–17]. This can only be guided by accurate data on its incidence in the ACL-injured knee.

The aim of this study was to determine the incidence and MRI characteristics of the spectrum of PLC injuries occurring in association with ACL rupture. The hypothesis was that the prevalence of PLC injury that occurs in the ACL-injured knee is higher than previously reported.

Materials and methods

Institutional Review Board approval was granted for this study. All patients with a clinical diagnosis of ACL rupture undergoing MRI evaluation of the knee between 1 July 2015 and 30 June 2016 at first assessment were included. Informed consent was obtained from patients before enrolment in the study.

All patients presented with acute injuries and underwent MRI within 1 week of the initial clinical examination. In addition to standard MRI knee reporting practice, particular emphasis was placed on identifying injury to the PLC in its complexity (Fig. 1) and describing precisely the spectrum of involvement of these structures, as previously described [12, 16, 18–22]. Two independent radiologists with more than 10 years of experience in musculoskeletal radiology evaluated all cases. Previous radiological and anatomical descriptions were used as a basis for interpretation of the MR imaging of the PLC (Figs. 2, 3) [9, 11, 16, 17, 19, 20, 22–24].

Fig. 1
figure 1

Illustration demonstrating the complexity of the normal anatomy of the posterolateral corner. a Axial, b Sagittal view

Full size image
Fig. 2
figure 2

MRI images demonstrating the normal posterolateral corner. a Sagittal view with arcuate ligament, capsule, and popliteomeniscal fascicles. b Coronal view of biceps femoris tendon. c Coronal view of lateral collateral ligament. d Axial view of the popliteus tendon. e Axial view of lateral collateral ligament. f Coronal view of popliteofibular ligament. g Sagittal view of popliteus tendon

Full size image
Fig. 3
figure 3

MRI images demonstrating lesions in the posterolateral corner. a Sagittal view with arcuate ligament tear. b Sagittal view of popliteomeniscal fascicle tears. c Coronal view of lateral collateral ligament and popliteus tendon tear. d Coronal view of popliteofibular ligament tear. e Axial view of popliteus tendon tear. f Axial view of arcuate ligament tear

Full size image

Magnetic resonance imaging was performed using a 1.5-T magnet with a wide-bore configuration (MAGNETOM Avanto, Siemens, Munich, Germany). Each scanning protocol commenced with a scout image in the axial, sagittal, and coronal planes. The axial images obtained were true axials and were derived directly from the scout image. In this image, injuries to the lateral meniscus, lateral collateral ligament (LCL), posterolateral joint capsule, and arcuate and fabellofibular ligaments were analyzed. The sagittal images were based on anatomical landmarks to individualize the sequence, ensuring proper visualization of the ACL and its two bundles for every patient. A plane was prescribed along the lateral femoral epicondyle at the level of the LCL, which is a constant anatomical landmark with little interpersonal variability. Evaluation of injuries (sprain or rupture) to the popliteus tendon, popliteomeniscal fascicles (superior and inferior), LCL, popliteofibular ligament (PFL), biceps tendon, and posterolateral joint capsule were made. An oblique sagittal sequence was also obtained. This plane was parallel to the ACL and allowed optimal visualization of both the anteromedial (AM) and posterolateral (PL) bundles. Routine coronal images for LCL, PFL, and biceps tendon evaluation and then special oblique–coronal images were obtained in the long axis of the ACL, starting at the intercondylar roof of Blumensaat’s line. The coronal–oblique sequence increases the sensitivity and specificity of diagnosing isolated AM or PL bundle injuries. It also helps to visualize the proximal insertion of the bundles for hemorrhage and rupture. A slice thickness of 3 mm can differentiate the two bundles as separate entities. In imaging these patients, it was important that the sagittal images covered the lateral-most portion of the knee joint so that the meniscocapsular junction was noted [19, 21].

The diagnosis of meniscal tears was performed using widely accepted criteria and reported according to their location [25, 26]. Injuries to the medial and LCLs were graded (0–3) according to Schweitzer et al., [27], and all injuries were considered significant for the purposes of this study according to Pacheco et al. [11]. For the PLC injuries, if the contour of the structures analyzed was irregular or if ligamentous edema existed, then the radiologists considered the structure to be abnormal. If only periligamentous edema existed, with identifiable, continuous low-signal intensity fibers, the ligament was considered intact [28]. Osseous injuries including bone contusions, cortical depression and trabecular fractures were located on the lateral and medial femoral condyle in addition to the lateral and medial tibial plateau. Injuries were classified as anterior, central, and posterior and were counted separately [29, 30].

All analyses were made with using the SPSS software (version 20.0, SPSS, Chicago, IL, USA) and the significance level was set at 5%. The Chi-squared test was used to evaluate the association between PLC and other injuries. Spearman’s rank correlation coefficient and logistic regression were also used to determine whether any variables were predictive of concomitant PLC injury in the ACL-deficient knee. Kappa (κ) values were measured to assess inter- and intra-reader agreement for determining the PLC lesions on MRI. The κ values were interpreted according to Landis and Koch recommendations [31].

Results

One hundred and sixty-two patients had a clinical diagnosis of ACL injury and subsequent MRI evaluation of the knee during the study period. MRI-proven combined ACL and PLC injuries occurred in 32 patients (19.7%). These patients were then further evaluated. Twenty-six patients (81.3%) were male. The mean of age was 32 years (±13). Right- and left-sided injuries occurred with equal frequency. Bone contusions occurred with the following frequencies: lateral tibial plateau 75%, lateral femoral condyle 75%, medial tibial plateau 46.9%, and medial femoral condyle 28.1%. Of the 32 patients, 46.9% had an MRI-proven medial meniscus lesion and 65.6% had a lesion of the lateral meniscus.

The distribution of the frequency of specific lesions involving the PLC is reported in Fig. 4. Table 1 demonstrates the correlation between concomitant injuries and the four most common structures of the PLC involved in this series (inferior and superior popliteomeniscal fascicle, popliteus tendon, and arcuate ligament). The presence of bone contusions was strongly correlated with superior popliteomeniscal fascicle lesions (p < 0.05). The other structures of the PLC could not be statistically analyzed for correlation with concomitant injuries owing to insufficient numbers. However, there was no correlation between other lesions of the PLC and other lesions after logistic regression analysis (p > 0.05). Intra-reader agreement was good for both readers (κ = 0.79 and 0.86). There was also good inter-reader agreement (κ = 0.79).

Fig. 4
figure 4

Bar graph showing the prevalence of each specific posterolateral corner lesion (N = 32)

Full size image
Table 1 Correlation findings at posterolateral corner lesions and every other anatomical site injury
Full size table

Discussion

In this MRI-based study, the main finding was that the incidence of PLC injury in the ACL-deficient knee was 19.7%. This is higher than the previously reported rates of 5–14.7% [11, 13, 14, 23, 24, 32, 33]. It is postulated that the relatively high rate of concomitant injuries to the PLC reported in the current series is a reflection of previous under-reporting and missed diagnoses, as alluded to by several authors [11, 13, 14, 23, 24, 32].

Although PLC injury is a well-known entity, it is surprisingly frequently missed, probably because of a poor understanding of the complex anatomy of this region [11, 19, 32, 34]. The major confusion about PLC anatomy found in the orthopedic literature lies with the structures that course from the fibular head and that attach to the popliteus complex and the joint capsule of the PL aspect of the knee. All these structures are difficult to identify separately in cadaveric and clinical evaluations. This anatomical complexity explains why most studies describe only the main structures of the PLC (LCL, PFL, and popliteus tendon) and this contributes to the relative under-diagnosis of injuries in previously reported studies compared with the current study, which evaluates all the structures within the PLC [15, 17, 19, 35]. In addition, in contrast to previous studies that evaluated patients with grade 3 PLC injury, the current study also included patients with grade 1 and 2 injuries. However, there are a number of additional reasons for the under-diagnosis of PLC injury. These include failure to perform appropriate clinical tests or using tests which have a low sensitivity [11, 14, 17, 18, 36–39] and the timing of imaging studies [11, 14, 17, 18, 36]. It is recognized that MRI performed within 3 weeks of the initial knee trauma is associated with better sensitivity and specificity in identifying PLC lesions than delayed imaging [11, 14, 18, 19, 21].

It is apparent that a greater awareness of the normal and abnormal MRI appearances of the structures of the PLC of the knee and of the patterns of injury often seen in patients with PLC rotatory instability will help clinicians suggest the diagnosis of PLC injury, even if it is not clinically detected [16, 19, 24]. The fact that the rate of combined ACL and PLC injuries in this series was almost 20%, supports the role of thorough clinical and MRI evaluation for PLC injury in all ACL-deficient knees [11, 18, 36]. Although many articles have reported the role of the PLC in rotatory stability, to our knowledge, this is the first to report a detailed radiological and anatomical characterization of the spectrum of injury in a large prospective series of ACL-injured knees [8, 33, 35, 37, 40, 41].

In this study, bone contusions occurred in 75% of patients in the lateral tibial plateau and lateral femoral condyle. The rate of bone contusions in patients with an ACL rupture but without a PLC injury was not evaluated, as the purpose of this series was to evaluate only the spectrum of PLC injuries occurring in association with ACL rupture. There was a statistically significant correlation between lateral bone contusions and superior popliteomeniscal fascicle lesions. However, this pattern of bone bruising is not specific to this injury, as other authors have also reported that lateral bone contusions are associated with lateral meniscal lesions, MCL tear, anterolateral ligament injury, high-grade (grade II and III) pivot shift, and that the severity of these injuries correlates with the degree of bone contusion [15, 30, 42]. Yoon et al. reported the same pattern, where the incidence of injuries to the menisci and the MCL were positively associated with the incidence of bone contusion [42]. In addition, Shaw et al. reported that lateral meniscus injuries were associated with a higher incidence of PLC injury (p > 0.05) [15]. These findings suggest that where any features of increased severity of knee injury might exist, the index of suspicion of injury to the PLC should be raised and a more careful clinical and radiological evaluation of those structures is warranted.

The main limitation of this study is the relatively small number of patients with PLC injury included; however, this reflects previous literature on this topic and includes all patients seen within a 12-month period in a busy tertiary referral specialist knee clinic. Despite the small numbers, the lack of similar studies characterizing the spectrum of PLC injuries in ACL-deficient knees means that this study is able to provide useful benchmark data to highlight the frequency and extent of these combined injuries. However, it is recognized that studies with larger numbers of patients may further elucidate the complex relationships of this lesion with ACL injuries and other lesions. A further important limitation is the absence of clinical follow-up of these patients; in particular, it is not known whether they required PLC reconstruction. The lack of these data means that it is not possible to correlate the radiological severity of PLC injury with clinically apparent instability. However, previous authors have suggested that the presence of tears of two or more structures of the PLC is a hallmark of high-grade injury and should direct the orthopedic surgeon to carefully examine the rotatory instability and evaluate if repair or reconstruction are necessary [7, 8, 16, 20, 22, 34, 37, 43]. Another limitation is that we did not record the precise mechanism of injury. It may be the case that the rate of injury to the PLC identified in this series is higher than in previous reports because of the heterogeneity with regard to mechanisms of injury. It is also possible that these findings were influenced by observer bias, as the radiologists were not blinded. Therefore, there was a risk of overcalling injuries, particularly when the focus of the study was to evaluate injuries to the PLC. However, the study design attempted to minimize bias by an independent evaluation of imaging by two highly experienced radiologists. Good inter- and intra-reader agreement provides some reassurance that the risk of the influence of observer bias on the key findings of this study was low.

The key message from this study is that PLC injury is more common than previously reported. This should result in an increased index of suspicion when evaluating the ACL-injured knee. Clearly, further studies will be required to clarify the long-term clinical relevance of MRI observations [8].

Conclusion

The incidence of PLC injury in the ACL-deficient knee has previously been under-reported. This series demonstrates that nearly 20% of patients with an ACL rupture have some injury to the PLC when evaluated by MRI. Further clinical and biomechanical studies are required to clarify the long-term clinical relevance of these injuries.